Record the provider health validation and add the changelog entry

The user ran the real-hypervisor checks against beelink and confirmed they
passed, which completes Step 7 of the provider health plan.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

CHANGELOG.md

@@ -1,5 +1,76 @@
 # Changelog
 
+## 2026-09-16 08:45 UTC — Provider health checks: ProviderHealth CRD, per-provider checks, gauges
+
+- **Each configured provider is now checked on a schedule** and reported in a
+  cluster-scoped `ProviderHealth` object (`kubectl get ph`), one per provider, with a
+  `Ready` condition, a reason, and per-target verdicts.
+- **Checks:** kubernetes lists one object from the API server; GCP fetches a single
+  instance page; KVM checks SSH/libvirt reachability and the storage pool on every
+  host, in parallel. Failures map to fixed reasons (`Unreachable`,
+  `SSHHandshakeFailed`, `PoolNotFound`, `Timeout`, …).
+- **The KVM check can't hurt reconciles:** it uses `hostPool.probe`, which gives up on
+  a slow RPC without dropping the shared connection, and waiting for a host slot can
+  be interrupted.
+- **Metrics:** `proxy_operator_provider_target_up`, `_check_duration_seconds` and
+  `_check_timestamp_seconds`. The flags are `--provider-check-interval` and
+  `--provider-check-timeout` (defaults to `min(15s, interval)`).
+- Validated with `make test`, `make test-e2e` and the real-hypervisor KVM run against
+  beelink. `make lint` is blocked by a Go 1.27 toolchain/staticcheck panic that has
+  nothing to do with this change.
+- Key files: `api/v1alpha1/providerhealth_types.go`, `internal/provider/check.go`,
+  `internal/provider/kvm/`, `internal/providerhealth/`, `docs/architecture.md`.
+
+## 2026-09-10 22:58 UTC — Code-review fixes on the KVM branch: deletion, validation, lazy SSH, guest agent
+
+Five findings from a review of `feat/kvm-provider`, all fixed on the branch.
+
+- **A Proxy can always finish deleting.** The delete path used to drop the finalizer
+  only when `Get` answered not-found, so a permanent error there retried forever —
+  a KVM host removed from the config was enough. It now calls `Delete` anyway, since
+  that is idempotent and can finish where `Get` cannot, and
+  `crawl.example.com/skip-cleanup: "true"` abandons the instance for whatever stays
+  stuck. In the KVM provider, deleting over an unconfigured host is a no-op success;
+  `Get` still fails permanently there so a typo cannot look like a missing VM.
+- **The guest agent ships as a vendor-data script**, superseding the cloud-config
+  merge in the entry below. Merging could never be made safe: cloud-init drops
+  vendor-data keys the user-data defines, and the strategy cannot be overridden from
+  either document. The Proxy's cloud-init now reaches the seed byte for byte in any
+  format — Jinja, MIME, gzip, scripts — where the round-trip previously re-typed
+  unquoted scalars (`version: 1.10` arrived as `1.1`), dropped comments and reordered
+  keys. New `installGuestAgent` setting, default on.
+- **A missing `kvm-ssh` Secret no longer crash-loops the manager.** The SSH material
+  is read on first use and cached, so only KVM Proxies fail and they recover when the
+  Secret appears. A missing file is transient, an unparseable key permanent, and
+  startup still logs a warning.
+- **Host names and image keys are validated.** Either can end up in a providerID or a
+  volume file name, where a slash made the VM unaddressable for good.
+- Verified end to end: unit suite, the integration test against a real hypervisor
+  (22 s, agent reporting the IP), both in-cluster Proxies Ready in 2 min with 204
+  through squid, the guest showing `qemu-guest-agent` active with its cloud-config
+  comments intact, and the manager staying up with the Secret pointed away.
+- Key files: `internal/provider/kvm/{seed.go,conn.go,kvm.go}`,
+  `internal/controller/proxy_controller.go`, `internal/provider/config_kvm.go`,
+  `api/v1alpha1/proxy_types.go`.
+
+## 2026-09-10 21:04 UTC — KVM provider passes in-cluster; guest agent merged into cloud-config
+
+- Two KVM Proxies with real cloud-configs reach `Ready` from inside the kind cluster
+  (2 min from apply, IP via the guest agent, 204 through squid from the LAN).
+- The `qemu-guest-agent` bootstrap is now merged into each Proxy's cloud-config
+  (`packages` + `runcmd`) instead of shipped as vendor-data: cloud-init's default
+  merger drops vendor-data keys the user-data also defines, so any Proxy with its own
+  `packages:` never got the agent and stayed `Provisioning` forever. `vendorData` is
+  now plain extra vendor-data with no default; non-cloud-config user-data (scripts,
+  `#include`) still gets the bootstrap as vendor-data. A cloud-config that is not a
+  YAML mapping fails the Proxy permanently instead of booting without its config.
+- The manager pod mounts a 64 MiB `emptyDir` on `/tmp`: the seed ISO writer stages
+  files on disk and the root filesystem is read-only.
+- Errors already classified with a provider sentinel no longer drop the cached
+  libvirt SSH connection; only libvirt and transport errors do.
+- Key files: `internal/provider/kvm/{seed.go,conn.go,errors.go}`,
+  `config/manager/manager.yaml`, `docs/kvm-provider.md`.
+
 ## 2026-09-10 10:07 UTC — KVM provider validated on a real hypervisor; UEFI default; libvirt call tracing
 
 - `make test-integration-kvm` passes end to end against a libvirt 11.6 / QEMU 10.1

CLAUDE.md

@@ -25,10 +25,11 @@ entrypoint yet, so there's nothing to run locally.
 
 ## Architecture
 
-Not yet built. TODO: once the controller-runtime manager and reconcile loop exist,
-document the data-flow diagram and package/role table here, following the standard
-reconcile-loop shape: fetch → resolve → list related → compute desired state → merge
-→ patch only on diff.
+Documented in [docs/architecture.md](docs/architecture.md). It holds the component
+table, the reconcile flow, and one section per runnable: health engine, discovery and
+lease API, orphan GC, metrics, tracing, and provider health checks. Provider-specific
+detail is in `docs/kvm-provider.md` and the GCP docs; the test layout is in
+`docs/testing.md`.
 
 ## Conventions
 
@@ -214,7 +215,9 @@ A lightweight `check` job (`go vet` / `go build` / `go test -short`) gates the b
 
 ## Gotchas
 
-- This repo currently contains only `go.mod` (module
-  `gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator`) — no controller code,
-  `main.go`, Dockerfile, k8s manifests, Makefile, or CI config exist yet. Nothing in
-  this file describes working software; it's the convention scaffold to build against.
+- The "What this is" and "Commands" sections above predate the code and are stale.
+  The Makefile targets (`make test`, `make manifests generate`, `make test-e2e`,
+  `make test-integration-kvm`) and `docs/testing.md` are authoritative.
+- `+kubebuilder:rbac` markers are package-level. Keep them in their own comment block,
+  never inside a type's or function's doc comment: controller-gen silently ignores
+  them there. After `make manifests`, check `git diff config/rbac/role.yaml`.

PROJECT

@@ -18,4 +18,11 @@ resources:
   kind: Proxy
   path: gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/api/v1alpha1
   version: v1alpha1
+- api:
+    crdVersion: v1
+  domain: example.com
+  group: crawl
+  kind: ProviderHealth
+  path: gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/api/v1alpha1
+  version: v1alpha1
 version: "3"

README.md

@@ -164,6 +164,12 @@ Cloud-init from a Secret: the Secret **must** carry the label
 labelled Secrets, so an unlabelled one is invisible (the Proxy reports
 `CloudInitError`). Rotating the Secret's content triggers VM replacement.
 
+Stuck deletion: a Proxy stays `Terminating` while the operator keeps failing
+to clean its instance up — credentials revoked, a provider or a KVM host
+dropped from the config. Annotate it with
+`crawl.example.com/skip-cleanup: "true"` to abandon the instance and release
+the object. Whatever is still running then has to be removed by hand.
+
 ## Providers
 
 `Managed` proxies are provisioned by a provider — a small compute backend

api/v1alpha1/providerhealth_types.go

@@ -0,0 +1,127 @@
+/*
+Copyright 2026.
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+*/
+
+package v1alpha1
+
+import (
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+)
+
+// ConditionReady summarizes a ProviderHealth's targets: True when every
+// target passed its last check.
+const ConditionReady = "Ready"
+
+// ProviderTarget is the latest check verdict for one thing a provider
+// depends on: a KVM hypervisor, the Kubernetes API server, a GCP project.
+type ProviderTarget struct {
+	// name identifies the target within its provider: the KVM host name,
+	// "apiserver", or the GCP project.
+	// +kubebuilder:validation:MinLength=1
+	// +required
+	Name string `json:"name"`
+
+	// address is where the check connected to.
+	// +optional
+	Address string `json:"address,omitempty"`
+
+	// ready is true when the last check succeeded.
+	// +required
+	Ready bool `json:"ready"`
+
+	// reason is a CamelCase verdict such as Reachable, Unreachable,
+	// Forbidden, SSHHandshakeFailed or PoolNotActive.
+	// +optional
+	Reason string `json:"reason,omitempty"`
+
+	// message is human-readable detail, usually the error of a failed check.
+	// +kubebuilder:validation:MaxLength=1024
+	// +optional
+	Message string `json:"message,omitempty"`
+
+	// latencyMillis is how long the last check took.
+	// +optional
+	LatencyMillis int64 `json:"latencyMillis,omitempty"`
+
+	// lastTransitionTime is when ready last changed.
+	// +required
+	LastTransitionTime metav1.Time `json:"lastTransitionTime"`
+
+	// details are provider-specific facts from the last check, e.g. the
+	// libvirt version and free storage-pool space of a KVM host.
+	// +optional
+	Details map[string]string `json:"details,omitempty"`
+}
+
+// ProviderHealthStatus is the result of the latest check of one provider.
+type ProviderHealthStatus struct {
+	// type is the provider's type from the providers config: kubernetes,
+	// gcp or kvm.
+	// +optional
+	Type string `json:"type,omitempty"`
+
+	// conditions hold Ready, which summarizes the targets.
+	// +listType=map
+	// +listMapKey=type
+	// +optional
+	Conditions []metav1.Condition `json:"conditions,omitempty"`
+
+	// targets hold one verdict per checked target.
+	// +listType=map
+	// +listMapKey=name
+	// +optional
+	Targets []ProviderTarget `json:"targets,omitempty"`
+
+	// targetsReady counts the ready targets as "ready/total".
+	// +optional
+	TargetsReady string `json:"targetsReady,omitempty"`
+
+	// lastCheckTime is when the latest check finished.
+	// +optional
+	LastCheckTime *metav1.Time `json:"lastCheckTime,omitempty"`
+}
+
+// +kubebuilder:object:root=true
+// +kubebuilder:subresource:status
+// +kubebuilder:resource:scope=Cluster,shortName=ph
+// +kubebuilder:printcolumn:name="Type",type=string,JSONPath=`.status.type`
+// +kubebuilder:printcolumn:name="Ready",type=string,JSONPath=`.status.conditions[?(@.type=="Ready")].status`
+// +kubebuilder:printcolumn:name="Targets",type=string,JSONPath=`.status.targetsReady`
+// +kubebuilder:printcolumn:name="Last Check",type=date,JSONPath=`.status.lastCheckTime`
+// +kubebuilder:printcolumn:name="Age",type=date,JSONPath=`.metadata.creationTimestamp`
+
+// ProviderHealth reports the periodic health check of one configured
+// provider. It is named after the provider, and the operator creates,
+// updates and deletes it itself; there is no spec to edit.
+type ProviderHealth struct {
+	metav1.TypeMeta `json:",inline"`
+
+	// metadata is a standard object metadata
+	// +optional
+	metav1.ObjectMeta `json:"metadata,omitzero"`
+
+	// status is the result of the latest check
+	// +optional
+	Status ProviderHealthStatus `json:"status,omitzero"`
+}
+
+// +kubebuilder:object:root=true
+
+// ProviderHealthList contains a list of ProviderHealth
+type ProviderHealthList struct {
+	metav1.TypeMeta `json:",inline"`
+	metav1.ListMeta `json:"metadata,omitzero"`
+	Items           []ProviderHealth `json:"items"`
+}

api/v1alpha1/proxy_types.go

@@ -62,6 +62,15 @@ const (
 	// must be replaced.
 	AnnotationSpecHash = "crawl.example.com/spec-hash"
 
+	// AnnotationSkipCleanup abandons the provider-side instance: set it to
+	// "true" and deletion removes the finalizer without asking the provider
+	// to clean anything up. It is the escape hatch for a cleanup that cannot
+	// succeed — a provider dropped from the config, credentials revoked, a
+	// backend that answers permanently — where the object would otherwise
+	// stay in Terminating forever. The instance is left running and becomes
+	// the administrator's to remove.
+	AnnotationSkipCleanup = "crawl.example.com/skip-cleanup"
+
 	// LabelCloudInit must be set (to "true") on every Secret referenced by
 	// spec.cloudInit.secretRef: the manager's cache only holds Secrets
 	// carrying this label, so an unlabelled Secret is invisible to the
@@ -352,7 +361,7 @@ type ProxyList struct {
 
 func init() {
 	SchemeBuilder.Register(func(s *runtime.Scheme) error {
-		s.AddKnownTypes(SchemeGroupVersion, &Proxy{}, &ProxyList{})
+		s.AddKnownTypes(SchemeGroupVersion, &Proxy{}, &ProxyList{}, &ProviderHealth{}, &ProviderHealthList{})
 		return nil
 	})
 }

api/v1alpha1/zz_generated.deepcopy.go

@@ -95,6 +95,120 @@ func (in *PlacementSpec) DeepCopy() *PlacementSpec {
 	return out
 }
 
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ProviderHealth) DeepCopyInto(out *ProviderHealth) {
+	*out = *in
+	out.TypeMeta = in.TypeMeta
+	in.ObjectMeta.DeepCopyInto(&out.ObjectMeta)
+	in.Status.DeepCopyInto(&out.Status)
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ProviderHealth.
+func (in *ProviderHealth) DeepCopy() *ProviderHealth {
+	if in == nil {
+		return nil
+	}
+	out := new(ProviderHealth)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.
+func (in *ProviderHealth) DeepCopyObject() runtime.Object {
+	if c := in.DeepCopy(); c != nil {
+		return c
+	}
+	return nil
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ProviderHealthList) DeepCopyInto(out *ProviderHealthList) {
+	*out = *in
+	out.TypeMeta = in.TypeMeta
+	in.ListMeta.DeepCopyInto(&out.ListMeta)
+	if in.Items != nil {
+		in, out := &in.Items, &out.Items
+		*out = make([]ProviderHealth, len(*in))
+		for i := range *in {
+			(*in)[i].DeepCopyInto(&(*out)[i])
+		}
+	}
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ProviderHealthList.
+func (in *ProviderHealthList) DeepCopy() *ProviderHealthList {
+	if in == nil {
+		return nil
+	}
+	out := new(ProviderHealthList)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.
+func (in *ProviderHealthList) DeepCopyObject() runtime.Object {
+	if c := in.DeepCopy(); c != nil {
+		return c
+	}
+	return nil
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ProviderHealthStatus) DeepCopyInto(out *ProviderHealthStatus) {
+	*out = *in
+	if in.Conditions != nil {
+		in, out := &in.Conditions, &out.Conditions
+		*out = make([]v1.Condition, len(*in))
+		for i := range *in {
+			(*in)[i].DeepCopyInto(&(*out)[i])
+		}
+	}
+	if in.Targets != nil {
+		in, out := &in.Targets, &out.Targets
+		*out = make([]ProviderTarget, len(*in))
+		for i := range *in {
+			(*in)[i].DeepCopyInto(&(*out)[i])
+		}
+	}
+	if in.LastCheckTime != nil {
+		in, out := &in.LastCheckTime, &out.LastCheckTime
+		*out = (*in).DeepCopy()
+	}
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ProviderHealthStatus.
+func (in *ProviderHealthStatus) DeepCopy() *ProviderHealthStatus {
+	if in == nil {
+		return nil
+	}
+	out := new(ProviderHealthStatus)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ProviderTarget) DeepCopyInto(out *ProviderTarget) {
+	*out = *in
+	in.LastTransitionTime.DeepCopyInto(&out.LastTransitionTime)
+	if in.Details != nil {
+		in, out := &in.Details, &out.Details
+		*out = make(map[string]string, len(*in))
+		for key, val := range *in {
+			(*out)[key] = val
+		}
+	}
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ProviderTarget.
+func (in *ProviderTarget) DeepCopy() *ProviderTarget {
+	if in == nil {
+		return nil
+	}
+	out := new(ProviderTarget)
+	in.DeepCopyInto(out)
+	return out
+}
+
 // DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
 func (in *Proxy) DeepCopyInto(out *Proxy) {
 	*out = *in

cmd/main.go

@@ -16,8 +16,8 @@ limitations under the License.
 
 // Package main is the composition root: it loads the provider config (fail
 // fast), assembles the provider registry, and wires the reconciler, health
-// engine, lease store, discovery API, orphan GC, and metrics onto one
-// controller-runtime manager.
+// engine, lease store, discovery API, orphan GC, provider health checks and
+// metrics onto one controller-runtime manager.
 package main
 
 import (
@@ -60,6 +60,7 @@ import (
 	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider/kubernetes"
 	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider/kvm"
 	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider/registry"
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/providerhealth"
 	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/tracing"
 	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/version"
 	// +kubebuilder:scaffold:imports
@@ -97,6 +98,7 @@ func main() {
 	var showVersion bool
 	var gcpWireFullPayloads bool
 	var traceHealthProbes bool
+	var providerCheckInterval, providerCheckTimeout time.Duration
 
 	flag.StringVar(&metricsAddr, "metrics-bind-address", "0", "The address the metrics endpoint binds to. "+
 		"Use :8443 for HTTPS or :8080 for HTTP, or leave as 0 to disable the metrics service.")
@@ -140,6 +142,13 @@ func main() {
 	flag.BoolVar(&traceHealthProbes, "trace-health-probes", false,
 		"Emit one trace span per health probe. Off by default: probes run about "+
 			"once per second per proxy and would dominate trace volume.")
+	flag.DurationVar(&providerCheckInterval, "provider-check-interval", time.Minute,
+		"Interval between provider health checks (can the cluster reach each hypervisor, API "+
+			"and account), published as ProviderHealth objects and metrics. 0 turns them off; "+
+			"otherwise at least 10s.")
+	flag.DurationVar(&providerCheckTimeout, "provider-check-timeout", 0,
+		"Time limit for one round of provider health checks: 15s, or the interval when that is "+
+			"shorter. If set, at most the interval.")
 
 	opts := zap.Options{
 		Development: true,
@@ -174,6 +183,10 @@ func main() {
 		setupLog.Error(nil, "--providers-config is required")
 		os.Exit(1)
 	}
+	if err := providerhealth.ValidateSchedule(providerCheckInterval, providerCheckTimeout); err != nil {
+		setupLog.Error(err, "Invalid provider check flags")
+		os.Exit(1)
+	}
 	cfg, err := provider.LoadConfigFile(providersConfig)
 	if err != nil {
 		setupLog.Error(err, "Failed to load providers config", "path", providersConfig)
@@ -195,6 +208,14 @@ func main() {
 		os.Exit(1)
 	}
 	m := metrics.New()
+	// Checkers come from the unwrapped providers, in config order: the
+	// metrics and tracing wrappers implement only Provider, and a check is
+	// not a provider request to be counted anyway.
+	checkTargets := make([]providerhealth.Target, 0, len(cfg.Providers))
+	for _, pc := range cfg.Providers {
+		checker, _ := providers[pc.Name].(provider.Checker)
+		checkTargets = append(checkTargets, providerhealth.Target{Name: pc.Name, Type: pc.Type, Checker: checker})
+	}
 	for name, p := range providers {
 		// Tracing outermost: the span covers the metrics recording too.
 		providers[name] = provider.WithTracing(name, provider.WithMetrics(name, p, m))
@@ -314,6 +335,18 @@ func main() {
 		os.Exit(1)
 	}
 
+	monitor := &providerhealth.Monitor{
+		Client:   mgr.GetClient(),
+		Reader:   mgr.GetAPIReader(),
+		Targets:  checkTargets,
+		Interval: providerCheckInterval,
+		Timeout:  providerCheckTimeout,
+	}
+	if err := mgr.Add(monitor); err != nil {
+		setupLog.Error(err, "Failed to add provider health monitor")
+		os.Exit(1)
+	}
+
 	if err := m.Register(ctrlmetrics.Registry,
 		proxyPhaseCounts(mgr.GetClient()),
 		func() int {
@@ -327,6 +360,10 @@ func main() {
 		setupLog.Error(err, "Failed to register metrics")
 		os.Exit(1)
 	}
+	if err := metrics.RegisterProviderTargets(ctrlmetrics.Registry, providerTargetSamples(monitor)); err != nil {
+		setupLog.Error(err, "Failed to register provider check metrics")
+		os.Exit(1)
+	}
 
 	if err := (&controller.ProxyReconciler{
 		Client:       mgr.GetClient(),
@@ -389,3 +426,23 @@ func proxyPhaseCounts(c client.Reader) func() map[string]int {
 		return counts
 	}
 }
+
+// providerTargetSamples flattens the monitor's latest results for the
+// scrape-time provider-check gauges.
+func providerTargetSamples(mon *providerhealth.Monitor) func() []metrics.ProviderTarget {
+	return func() []metrics.ProviderTarget {
+		var out []metrics.ProviderTarget
+		for _, r := range mon.Snapshot() {
+			for _, ts := range r.Targets {
+				out = append(out, metrics.ProviderTarget{
+					Provider:  r.Provider,
+					Target:    ts.Name,
+					Up:        ts.Ready,
+					Duration:  ts.Latency,
+					CheckedAt: r.CheckedAt,
+				})
+			}
+		}
+		return out
+	}
+}

config/crd/bases/crawl.example.com_providerhealths.yaml

@@ -0,0 +1,193 @@
+---
+apiVersion: apiextensions.k8s.io/v1
+kind: CustomResourceDefinition
+metadata:
+  annotations:
+    controller-gen.kubebuilder.io/version: v0.21.0
+  name: providerhealths.crawl.example.com
+spec:
+  group: crawl.example.com
+  names:
+    kind: ProviderHealth
+    listKind: ProviderHealthList
+    plural: providerhealths
+    shortNames:
+    - ph
+    singular: providerhealth
+  scope: Cluster
+  versions:
+  - additionalPrinterColumns:
+    - jsonPath: .status.type
+      name: Type
+      type: string
+    - jsonPath: .status.conditions[?(@.type=="Ready")].status
+      name: Ready
+      type: string
+    - jsonPath: .status.targetsReady
+      name: Targets
+      type: string
+    - jsonPath: .status.lastCheckTime
+      name: Last Check
+      type: date
+    - jsonPath: .metadata.creationTimestamp
+      name: Age
+      type: date
+    name: v1alpha1
+    schema:
+      openAPIV3Schema:
+        description: |-
+          ProviderHealth reports the periodic health check of one configured
+          provider. It is named after the provider, and the operator creates,
+          updates and deletes it itself; there is no spec to edit.
+        properties:
+          apiVersion:
+            description: |-
+              APIVersion defines the versioned schema of this representation of an object.
+              Servers should convert recognized schemas to the latest internal value, and
+              may reject unrecognized values.
+              More info: https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#resources
+            type: string
+          kind:
+            description: |-
+              Kind is a string value representing the REST resource this object represents.
+              Servers may infer this from the endpoint the client submits requests to.
+              Cannot be updated.
+              In CamelCase.
+              More info: https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#types-kinds
+            type: string
+          metadata:
+            type: object
+          status:
+            description: status is the result of the latest check
+            properties:
+              conditions:
+                description: conditions hold Ready, which summarizes the targets.
+                items:
+                  description: Condition contains details for one aspect of the current
+                    state of this API Resource.
+                  properties:
+                    lastTransitionTime:
+                      description: |-
+                        lastTransitionTime is the last time the condition transitioned from one status to another.
+                        This should be when the underlying condition changed.  If that is not known, then using the time when the API field changed is acceptable.
+                      format: date-time
+                      type: string
+                    message:
+                      description: |-
+                        message is a human readable message indicating details about the transition.
+                        This may be an empty string.
+                      maxLength: 32768
+                      type: string
+                    observedGeneration:
+                      description: |-
+                        observedGeneration represents the .metadata.generation that the condition was set based upon.
+                        For instance, if .metadata.generation is currently 12, but the .status.conditions[x].observedGeneration is 9, the condition is out of date
+                        with respect to the current state of the instance.
+                      format: int64
+                      minimum: 0
+                      type: integer
+                    reason:
+                      description: |-
+                        reason contains a programmatic identifier indicating the reason for the condition's last transition.
+                        Producers of specific condition types may define expected values and meanings for this field,
+                        and whether the values are considered a guaranteed API.
+                        The value should be a CamelCase string.
+                        This field may not be empty.
+                      maxLength: 1024
+                      minLength: 1
+                      pattern: ^[A-Za-z]([A-Za-z0-9_,:]*[A-Za-z0-9_])?$
+                      type: string
+                    status:
+                      description: status of the condition, one of True, False, Unknown.
+                      enum:
+                      - "True"
+                      - "False"
+                      - Unknown
+                      type: string
+                    type:
+                      description: type of condition in CamelCase or in foo.example.com/CamelCase.
+                      maxLength: 316
+                      pattern: ^([a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*/)?(([A-Za-z0-9][-A-Za-z0-9_.]*)?[A-Za-z0-9])$
+                      type: string
+                  required:
+                  - lastTransitionTime
+                  - message
+                  - reason
+                  - status
+                  - type
+                  type: object
+                type: array
+                x-kubernetes-list-map-keys:
+                - type
+                x-kubernetes-list-type: map
+              lastCheckTime:
+                description: lastCheckTime is when the latest check finished.
+                format: date-time
+                type: string
+              targets:
+                description: targets hold one verdict per checked target.
+                items:
+                  description: |-
+                    ProviderTarget is the latest check verdict for one thing a provider
+                    depends on: a KVM hypervisor, the Kubernetes API server, a GCP project.
+                  properties:
+                    address:
+                      description: address is where the check connected to.
+                      type: string
+                    details:
+                      additionalProperties:
+                        type: string
+                      description: |-
+                        details are provider-specific facts from the last check, e.g. the
+                        libvirt version and free storage-pool space of a KVM host.
+                      type: object
+                    lastTransitionTime:
+                      description: lastTransitionTime is when ready last changed.
+                      format: date-time
+                      type: string
+                    latencyMillis:
+                      description: latencyMillis is how long the last check took.
+                      format: int64
+                      type: integer
+                    message:
+                      description: message is human-readable detail, usually the error
+                        of a failed check.
+                      maxLength: 1024
+                      type: string
+                    name:
+                      description: |-
+                        name identifies the target within its provider: the KVM host name,
+                        "apiserver", or the GCP project.
+                      minLength: 1
+                      type: string
+                    ready:
+                      description: ready is true when the last check succeeded.
+                      type: boolean
+                    reason:
+                      description: |-
+                        reason is a CamelCase verdict such as Reachable, Unreachable,
+                        Forbidden, SSHHandshakeFailed or PoolNotActive.
+                      type: string
+                  required:
+                  - lastTransitionTime
+                  - name
+                  - ready
+                  type: object
+                type: array
+                x-kubernetes-list-map-keys:
+                - name
+                x-kubernetes-list-type: map
+              targetsReady:
+                description: targetsReady counts the ready targets as "ready/total".
+                type: string
+              type:
+                description: |-
+                  type is the provider's type from the providers config: kubernetes,
+                  gcp or kvm.
+                type: string
+            type: object
+        type: object
+    served: true
+    storage: true
+    subresources:
+      status: {}

config/crd/kustomization.yaml

@@ -3,4 +3,5 @@
 # It should be run by config/default
 resources:
 - bases/crawl.example.com_proxies.yaml
+- bases/crawl.example.com_providerhealths.yaml
 # +kubebuilder:scaffold:crdkustomizeresource

config/manager/kustomization.yaml

@@ -5,5 +5,5 @@ apiVersion: kustomize.config.k8s.io/v1beta1
 kind: Kustomization
 images:
 - name: controller
-  newName: example.com/egress-proxies-operator
-  newTag: v0.0.1
+  newName: egress-proxies-operator
+  newTag: kvm-dev

config/manager/manager.yaml

@@ -66,6 +66,10 @@ spec:
           - --providers-config=/etc/proxy-operator/providers.yaml
           # Add --trace-health-probes to emit one span per health probe
           # (high volume: ~1/s per proxy).
+          # Provider health checks (kubectl get providerhealth) run every
+          # minute; tune with --provider-check-interval (0 turns them off,
+          # otherwise at least 10s) and --provider-check-timeout (15s, or
+          # the interval if that is shorter).
         env:
           # Bearer token for the discovery API. Optional: without the
           # Secret the API serves unauthenticated (with a loud warning).
@@ -154,6 +158,11 @@ spec:
         - name: kvm-ssh
           mountPath: /etc/proxy-operator/kvm
           readOnly: true
+        # Scratch space: the root filesystem is read-only, and the kvm
+        # provider builds each VM's cloud-init seed ISO through a temp
+        # directory (a few hundred KB, deleted right after upload).
+        - name: tmp
+          mountPath: /tmp
       volumes:
       - name: providers-config
         configMap:
@@ -162,5 +171,8 @@ spec:
         secret:
           secretName: kvm-ssh
           optional: true
+      - name: tmp
+        emptyDir:
+          sizeLimit: 64Mi
       serviceAccountName: controller-manager
       terminationGracePeriodSeconds: 10

config/manager/providers_config.yaml

@@ -16,3 +16,39 @@ data:
     providers:
       - name: kubernetes
         type: kubernetes
+      - name: kvm-home
+        type: kvm
+        kvm:
+          # SSH to each host with the libvirt unix socket forwarded through the
+          # connection; nothing to open on the hypervisors beyond sshd. Both
+          # files come from the optional kvm-ssh Secret mounted by
+          # config/manager/manager.yaml — see docs/kvm-provider.md.
+          sshKeyFile: /etc/proxy-operator/kvm/id_ed25519
+          knownHostsFile: /etc/proxy-operator/kvm/known_hosts
+          # insecureIgnoreHostKey: true   # skip host-key checks (not recommended)
+          # diskSizeGb: 10                # per-VM overlay; cloud-init grows the root fs into it
+          # vendorData: |                 # extra vendor-data; the qemu-guest-agent bootstrap is merged
+          #   #cloud-config               # into each Proxy's cloud-config, not configured here
+          #   ...
+          images:
+            # placement.image names a key here (or is an https:// URL directly).
+            # Fetched into each host's pool on first use and cached there as
+            # base-<key>.qcow2; do not change an entry while Proxies use it.
+            debian-13:
+              url: https://cloud.debian.org/images/cloud/trixie/latest/debian-13-generic-amd64.qcow2
+              # sha256: "<64 hex>"        # optional; verified after download
+              # firmware: uefi            # default; "bios" only for images with no EFI partition
+              #                           # (Debian 13 hangs in GRUB under BIOS headless: Debian bug #1111240)
+            ubuntu-24.04:
+              url: https://cloud-images.ubuntu.com/noble/current/noble-server-cloudimg-amd64.img
+          machineTypes:                   # placement.machineType names a key here
+            small:  {vcpus: 1, memoryMiB: 1024}
+            medium: {vcpus: 2, memoryMiB: 2048}
+          hosts:                          # placement.zone names a host; optional with one host
+            - name: beelink
+              address: beelink.tail8c69c.ts.net:22
+              user: novakj                  # must be in the libvirt group
+              bridge: br0                 # to a LAN with DHCP; the VM's IP comes from there
+              # socket: /var/run/libvirt/libvirt-sock
+              # pool: default
+

config/rbac/kustomization.yaml

@@ -25,4 +25,7 @@ resources:
 - proxy_admin_role.yaml
 - proxy_editor_role.yaml
 - proxy_viewer_role.yaml
+# ProviderHealth objects are written only by the operator, so they get a
+# viewer role and nothing else.
+- providerhealth_viewer_role.yaml
 

config/rbac/providerhealth_viewer_role.yaml

@@ -0,0 +1,29 @@
+# This rule is not used by the project egress-proxies-operator itself.
+# It is provided to allow the cluster admin to help manage permissions for users.
+#
+# Grants read-only access to ProviderHealth objects, the results of the
+# operator's periodic provider checks. The operator creates, updates and
+# deletes them itself, so no admin or editor role is provided.
+
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRole
+metadata:
+  labels:
+    app.kubernetes.io/name: egress-proxies-operator
+    app.kubernetes.io/managed-by: kustomize
+  name: providerhealth-viewer-role
+rules:
+- apiGroups:
+  - crawl.example.com
+  resources:
+  - providerhealths
+  verbs:
+  - get
+  - list
+  - watch
+- apiGroups:
+  - crawl.example.com
+  resources:
+  - providerhealths/status
+  verbs:
+  - get

config/rbac/role.yaml

@@ -22,6 +22,22 @@ rules:
   - get
   - list
   - watch
+- apiGroups:
+  - crawl.example.com
+  resources:
+  - providerhealths
+  verbs:
+  - create
+  - delete
+  - get
+  - list
+- apiGroups:
+  - crawl.example.com
+  resources:
+  - providerhealths/status
+  verbs:
+  - get
+  - patch
 - apiGroups:
   - crawl.example.com
   resources:

config/samples/providers-config.yaml

@@ -30,8 +30,9 @@ providers:
       knownHostsFile: /etc/proxy-operator/kvm/known_hosts
       # insecureIgnoreHostKey: true   # skip host-key checks (not recommended)
       # diskSizeGb: 10                # per-VM overlay; cloud-init grows the root fs into it
-      # vendorData: |                 # default installs qemu-guest-agent for IP discovery
-      #   #cloud-config
+      # installGuestAgent: false      # skip the vendor-data script that installs qemu-guest-agent
+      # vendorData: |                 # extra vendor-data, carried beside the agent script as a
+      #   #cloud-config               # MIME multipart. The Proxy's own cloud-init is never rewritten.
       #   ...
       images:
         # placement.image names a key here (or is an https:// URL directly).
@@ -48,9 +49,9 @@ providers:
         small:  {vcpus: 1, memoryMiB: 1024}
         medium: {vcpus: 2, memoryMiB: 2048}
       hosts:                          # placement.zone names a host; optional with one host
-        - name: nas1
-          address: nas1.home.lan:22
-          user: virt                  # must be in the libvirt group
+        - name: beelink
+          address: beelink:22
+          user: novakj                  # must be in the libvirt group
           bridge: br0                 # to a LAN with DHCP; the VM's IP comes from there
           # socket: /var/run/libvirt/libvirt-sock
           # pool: default

docs/architecture.md

@@ -4,9 +4,9 @@
 
 | Component | Package | Runs as | Leader-elected | Role |
 |---|---|---|---|---|
-| Proxy CRD + helpers | `api/v1alpha1` | types | — | `Proxy` spec/status, CEL validation, defaulting, pure helpers |
+| Proxy + ProviderHealth CRDs | `api/v1alpha1` | types | — | `Proxy` spec/status, CEL validation, defaulting, pure helpers; the status-only, cluster-scoped `ProviderHealth` |
 | Reconciler | `internal/controller` | controller | yes (with the manager) | the state machine: provision, replace, delete, represent health |
-| Provider contract | `internal/provider` | library | — | `Provider` interface, error taxonomy, deterministic naming, config, metrics decorator |
+| Provider contract | `internal/provider` | library | — | `Provider` interface, optional `Checker`, error taxonomy, deterministic naming, config, metrics decorator |
 | kubernetes provider | `internal/provider/kubernetes` | library | — | real Squid pods in this cluster (local dev/CI) |
 | gcp provider | `internal/provider/gcp` | library | — | Compute Engine VMs, four API calls, fire-and-forget ops |
 | kvm provider | `internal/provider/kvm` | library | — | VMs on libvirt hosts over SSH-forwarded sockets; image cache, overlay disks, NoCloud seeds |
@@ -14,6 +14,7 @@
 | Lease store | `internal/lease` | Runnable (expiry sweep) | no | in-memory leases + cooldowns, single mutex |
 | Discovery API | `internal/discovery` | Runnable | no | HTTP list/lease/release/report on `:8090` |
 | Orphan GC | `internal/gc` | Runnable | yes | deletes tagged instances whose CR is gone |
+| Provider health monitor | `internal/providerhealth` | Runnable | yes | periodic provider checks → one `ProviderHealth` per provider, target gauges, transition logs |
 | Metrics | `internal/metrics` | library | — | explicit registration, scrape-time collectors |
 | Composition root | `cmd/main.go` | binary | — | flags, provider registry, wires everything onto one manager |
 
@@ -306,6 +307,8 @@ registries). Two kinds:
 - **Scrape-time collectors** — `proxy_operator_proxies{phase}` and
   `proxy_operator_leases_active` read the cache / lease store at every
   scrape; reconcile-incremented gauges inevitably drift and leak series.
+  The provider-check gauges (§11) read the health monitor's snapshot the
+  same way.
 - **Fed vectors** — `healthcheck_duration_seconds{proxy}` and
   `healthcheck_failures_total{proxy}` observe EVERY probe (status writes
   are transition-only; metrics carry the high-frequency signal), and the
@@ -348,7 +351,8 @@ incoming trace context):
 - Discovery API: one server span per request (named from the route
   pattern), incoming W3C `traceparent` honored so clients' traces continue
   into the operator; `/healthz` excluded.
-- `gc.sweep` per GC pass; `health.probe` per probe **only** with
+- `gc.sweep` per GC pass; `providerhealth.check` per round of provider
+  checks; `health.probe` per probe **only** with
   `--trace-health-probes` (default off — probes run ~1/s per proxy).
 
 Log correlation: logr sinks never see a context, so trace IDs ride on the
@@ -359,6 +363,63 @@ Re-deriving from the base rather than layering keeps zap from emitting
 duplicate keys on nested spans. GCP V(5) wire logs get the same keys from
 the slog handler's ctx.
 
+### 11. Provider health checks (`internal/providerhealth/`)
+
+Answers "can this cluster reach each provider's backend right now?"
+without waiting for a Proxy to fail. A provider opts in by implementing
+`provider.Checker`: `Check(ctx) []TargetStatus` returns one verdict per
+target and never fails as a whole. All three providers do:
+
+| Provider | Target | What one check does |
+|---|---|---|
+| kvm | each `kvm.hosts` entry | over the pooled SSH+libvirt connection: libvirt version and hostname, then the configured pool must exist and be running; reports pool capacity and free space ([details](kvm-provider.md#health-check)) |
+| kubernetes | `apiserver` | lists at most one managed Pod cluster-wide: the verb and scope `ListByTag` and GC rely on |
+| gcp | the project | one page of an aggregated instance list with `maxResults=1`: credentials, Compute API enabled, `compute.instances.list` |
+
+```text
+Monitor.Start(ctx)   ── leader-only; --provider-check-interval (1m; 0 = off,
+  │                     otherwise ≥ 10s), --provider-check-timeout (15s, or
+  │                     the interval if that is shorter)
+  ├─ once: delete managed-by-labelled ProviderHealth objects whose provider
+  │   is not configured (all of them when checks are off) — the providers
+  │   config is read once per process
+  └─ at start, then every Interval: checkOnce
+       │ every provider's Check concurrently, under one context.WithTimeout
+       │ snapshot → the scrape-time gauges
+       │ log transitions: down or reason changed → "WARNING: provider target
+       │   not ready"; back up → "Provider target ready"; steady → silent
+       └ per provider: APIReader.Get → Create if missing → status merge patch
+```
+
+- **One status write per provider per round**, since `lastCheckTime`
+  moves every round. At the 10s minimum interval that is still far below
+  leader-election renewals. Proxy status avoids writes like this only
+  because its probes run once a second per proxy.
+- **Reads go through the API reader**, not the cache. The objects are
+  cluster-scoped; a cached Get would start a cluster-wide informer (and
+  need `watch`) even with `--proxy-namespace`.
+- **Checks bypass the provider decorators.** `cmd/main.go` collects the
+  Checkers before wrapping the providers in `WithTracing(WithMetrics(...))`,
+  so a check never counts as a provider request. KVM RPCs still appear in
+  the V(1) `libvirt call` trace, and each round is one
+  `providerhealth.check` span.
+- **Not wired to `/readyz`.** A dead hypervisor must not make the manager
+  unready while the other providers keep working.
+- **Gauges** `proxy_operator_provider_target_up`,
+  `proxy_operator_provider_check_duration_seconds` and
+  `proxy_operator_provider_check_timestamp_seconds`, each labelled
+  `{provider,target}`. Only the leader exports them.
+
+Illustrative output:
+
+```sh
+$ kubectl get providerhealth          # short name: ph
+NAME         TYPE         READY   TARGETS   LAST CHECK   AGE
+kubernetes   kubernetes   True    1/1       12s          3d
+kvm-home     kvm          False   0/1       12s          3d
+$ kubectl get ph kvm-home -o yaml     # per target: reason, message, details
+```
+
 ## Decisions
 
 Judgment calls the spec left open, and deliberate deviations — recorded so
@@ -489,6 +550,15 @@ they read as choices, not accidents. Chronological by build step.
   cluster-wide Secret read RBAC — an unrestricted cache would hold every
   Secret in scope). Unlabelled referenced Secrets are invisible by
   construction, surfacing as `CloudInitError`.
+- **Deletion never gives up on its own, but can be released by hand:** a
+  permanent cleanup failure keeps retrying with the finalizer on, because
+  dropping it automatically would leak a running instance on any provider
+  that answers "permission denied". Where `Get` fails permanently the
+  reconciler still calls `Delete`, since it is idempotent and may finish
+  where `Get` cannot converge. `crawl.example.com/skip-cleanup: "true"`
+  abandons the instance and releases the object; it is checked before the
+  provider lookup, so it also frees a Proxy whose provider is gone from the
+  config.
 - **Events RBAC from the plan is omitted:** nothing wires an
   EventRecorder in the prototype, and granting verbs nothing uses would
   be RBAC lint noise. Add the marker together with the recorder if events
@@ -519,3 +589,27 @@ they read as choices, not accidents. Chronological by build step.
   external sites. The kube-apiserver ignores incoming `traceparent` by
   design (public endpoint), so k8s client spans are leaves — in-process
   traces, not cross-process ones.
+- **Provider health is a status-only, cluster-scoped CRD** (user
+  decision). `kubectl get providerhealth` works without Prometheus, which
+  the deployment does not ship (its ServiceMonitor is commented out). There
+  is no spec: the providers config file stays the one place providers are
+  configured, and the object only reports on it. The gauges carry the same
+  verdicts for alerting.
+- **KVM checks reuse the pooled connection instead of dialing fresh.** A
+  fresh dial would not pick up a rotated Secret anyway, since the SSH
+  material is cached. It would also open an sshd session per host per
+  minute. And a working cached connection is exactly what reconciles use.
+  Sharing must not cost the reconciles anything, so the check calls
+  through `hostPool.probe`. A check that runs out of time abandons its
+  RPCs instead of closing the connection, which would kill an upload or a
+  Create running on it. At most one abandoned check per host is left
+  waiting, and until it returns, the next check reports `Timeout` at once.
+  Waiting for a host's connection slot also ends with the caller's
+  deadline, so a reconcile redialing a dead host cannot stall the round.
+- **Provider transitions are logs, not Events**, in line with the
+  EventRecorder decision above. They log at Info with a `WARNING:` prefix,
+  and each target's `lastTransitionTime` in the object records when it
+  flipped.
+- **A provider name that is not a valid object name** gets metrics but no
+  `ProviderHealth` object, with a warning at start. Tightening config
+  validation instead would crash-loop existing deployments on upgrade.

docs/kvm-provider.md

@@ -27,11 +27,12 @@ Per host, once:
    `edk2-ovmf` on Fedora/RHEL. VMs boot UEFI by default (libvirt picks
    the OVMF binary itself); a domain start fails with "no firmware
    found" without it. Only `firmware: bios` images skip this.
-6. Optionally, **outbound internet from the VMs**: the default
-   vendor-data installs `qemu-guest-agent` from the distro mirrors on
-   first boot. Without it the VM still works — the operator falls back to
-   the host's ARP table to find the IP, which fills in once the guest
-   sends traffic.
+6. Optionally, **outbound internet from the VMs**: a vendor-data script
+   installs `qemu-guest-agent` from the distro mirrors on first boot.
+   Without it the VM still works — the operator falls back to the host's
+   ARP table to find the IP, which fills in once the guest sends traffic —
+   but the ARP table only sees traffic to the host itself, so on a plain
+   bridge the agent is in practice required.
 
 Check from the operator's point of view before wiring it in:
 
@@ -76,9 +77,10 @@ ssh virt@nas1.home.lan ip link show br0
    `/etc/proxy-operator/kvm` as two files, `id_ed25519` and
    `known_hosts`, which is what `sshKeyFile` and `knownHostsFile` in the
    provider config point at. The deployment starts fine without the
-   Secret as long as no `kvm` provider is configured. Rotating the key:
-   re-run the command (or the make target) and restart the manager; it
-   reads both files once at startup.
+   Secret, even with a `kvm` provider configured: only that provider's
+   proxies fail, and they recover once the Secret exists. Rotating the
+   key: re-run the command (or the make target) and restart the manager;
+   both files are read once, on the first connection, and cached.
 
 2. Add a provider entry to the providers config
    (`config/manager/providers_config.yaml` in-cluster; full sample in
@@ -102,9 +104,13 @@ ssh virt@nas1.home.lan ip link show br0
            bridge: br0
    ```
 
-   The operator reads the key and known_hosts at startup and refuses to
-   start if either is missing — a bad path fails fast, not on the first
-   Proxy.
+   The operator checks both files at startup and logs a `WARNING:` line
+   when either is unreadable, but it does not refuse to start: the
+   provider set is built all-or-nothing, so exiting here would crash-loop
+   the manager and take the `kubernetes` and `gcp` providers with it.
+   The files are read for real on the first connection. A missing file is
+   a transient failure, so proxies heal by themselves once the Secret
+   lands; a key that does not parse is permanent and needs a fix.
 
 3. Apply a Proxy (`config/samples/proxy_kvm.yaml`) and watch it:
 
@@ -129,13 +135,15 @@ ssh virt@nas1.home.lan ip link show br0
 | `knownHostsFile` | unless `insecureIgnoreHostKey` | | OpenSSH `known_hosts` to verify host keys against |
 | `insecureIgnoreHostKey` | | `false` | skip host-key verification entirely |
 | `diskSizeGb` | | `10` | capacity of each VM's overlay disk; cloud-init's growpart expands the root fs into it |
-| `vendorData` | | installs `qemu-guest-agent` | cloud-init vendor-data placed on every seed next to the Proxy's user-data |
+| `vendorData` | | | extra cloud-init vendor-data placed on every seed next to the Proxy's user-data. Beware that cloud-init drops any top-level key of a vendor-data **cloud-config** that the user-data also defines. Shipped as one part of a MIME multipart alongside the agent script when `installGuestAgent` is on |
+| `installGuestAgent` | | `true` | add the vendor-data script that installs and starts `qemu-guest-agent`. Turn off for images that ship it already or cannot reach a package mirror; IP discovery then depends on the host's ARP table |
+| `images.<key>` | | | the key names the image in `placement.image`. Letters, digits, dots, dashes and underscores only — it becomes the cached volume name `base-<key>.<format>` |
 | `images.<key>.url` | yes per entry | | http(s) URL of a cloud image (qcow2 or raw) |
 | `images.<key>.sha256` | | | expected digest; verified after download, a change triggers a re-download |
 | `images.<key>.format` | | `qcow2` | `qcow2` or `raw` |
 | `images.<key>.firmware` | | `uefi` | `uefi` (OVMF, secure boot off) or `bios` (SeaBIOS) for images without an EFI system partition; Debian 13 hangs in GRUB under BIOS on a headless VM, see *Known gaps* |
 | `machineTypes.<key>` | ≥ 1 | | `{vcpus, memoryMiB}` size profile |
-| `hosts[].name` | yes | | the `placement.zone` value that selects this host; unique |
+| `hosts[].name` | yes | | the `placement.zone` value that selects this host; unique. Letters, digits, dots, dashes and underscores only — it becomes part of the providerID `hosts/<name>/domains/<vm>` |
 | `hosts[].address` | yes | | SSH endpoint, `host:port` |
 | `hosts[].user` | yes | | SSH user, member of the `libvirt` group |
 | `hosts[].socket` | | `/var/run/libvirt/libvirt-sock` | libvirt socket path on the host |
@@ -174,6 +182,28 @@ console, the guest-agent channel, a virtio RNG and no graphics. Its
 creation time) — that is how orphan GC recognises its own VMs, since
 libvirt has no labels.
 
+The seed's `user-data` is the Proxy's `cloudInit`, **byte for byte**. Any
+format cloud-init accepts works, including a Jinja template, a MIME
+archive, a shell script or a gzipped payload; the operator never parses or
+re-renders it.
+
+The guest agent rides along in `vendor-data`, as a **shell script** rather
+than a cloud-config. That is the only shape cloud-init cannot drop:
+vendor-data cloud-config is merged under the user's own document with
+mergers that keep the user-data's value for every key both define and never
+append lists, so a Proxy carrying its own `packages:` silently lost the
+agent and sat in `Provisioning` forever. The strategy cannot be overridden
+from either document — `merge_how` in user-data governs only the first
+source, and `merge_how` in vendor-data is consumed and discarded before the
+document is stored. A vendor-data script skips all of it: cloud-init gives
+vendor-data its own handler chain, writes scripts to a separate directory
+and runs them from `cc_scripts_vendor`, after the user's packages are
+installed and before the user's `runcmd`. The script is idempotent, picks
+whichever of `apt-get`, `dnf`, `yum`, `zypper` or `apk` the image has, and
+logs into `/var/log/cloud-init-output.log`. Set `installGuestAgent: false`
+to leave it out. When `vendorData` is also configured, the two travel as
+separate parts of a MIME multipart vendor-data.
+
 ## The image cache
 
 The base image is downloaded **by the operator, through libvirt**, into
@@ -235,18 +265,79 @@ The provider logs through controller-runtime's logger, named `kvm`, with
   only `length` and `duration`. There is no raw wire tier like the GCP
   provider's V(5); go-libvirt exposes no hook for it.
 
+## Health check
+
+The operator checks every configured host every `--provider-check-interval`
+(default 1m). It publishes the result in two places: the `ProviderHealth`
+object named after the provider (`kubectl get ph kvm-home -o yaml`), and the
+`proxy_operator_provider_target_up{provider,target}` gauge.
+
+All hosts are checked concurrently, each over the same pooled connection
+that Create and Get use:
+
+1. The SSH and libvirt handshakes. These only happen on the first check, or
+   after the link dropped.
+2. `virConnectGetLibVersion` and `virConnectGetHostname`.
+3. `virStoragePoolLookupByName` and `virStoragePoolGetInfo` on the host's
+   `pool`.
+
+A passing target's `details` carry `hostname`, `libvirtVersion`, `pool`,
+`poolState`, `poolCapacity` and `poolAvailable` (in GiB). The free space is
+the closest thing this provider has to a capacity signal. Every target also
+shows its SSH `user`. A failing target says where the check broke:
+
+| Reason | Meaning | Look at |
+|---|---|---|
+| `SSHMaterialUnavailable` | the key or known_hosts file is missing or does not parse | the `kvm-ssh` Secret (see *Cluster-side setup*) |
+| `Unreachable` | the TCP connect failed, or the link died mid-check | DNS, Tailscale or a firewall between the cluster and `address` |
+| `SSHHandshakeFailed` | the host key is missing from known_hosts or has changed, or `user` rejected the key | the known_hosts entry; `user`'s `authorized_keys` |
+| `LibvirtConnectFailed` | SSH worked, the libvirt socket did not | `user` in the `libvirt` group, `libvirtd` running, the `socket` path |
+| `Timeout` | no answer within `--provider-check-timeout` (default 15s), or the previous check of the host is still waiting | a hung `libvirtd`, a saturated link, or a reconcile stuck redialing the host |
+| `PoolNotFound` | there is no pool named `pool` | `virsh pool-list --all` |
+| `PoolNotActive` | the pool exists but is not running | `virsh pool-start <pool>` and `virsh pool-autostart <pool>` |
+| `LibvirtError` | any other libvirt error | the message carries libvirt's own text |
+
+Sharing the connection has two consequences:
+
+- The leader now keeps one SSH connection per host open from startup, with a
+  keepalive every 15s. Before, the first connection opened with the first KVM
+  reconcile.
+- A check that runs out of time gives up on its RPCs but leaves the
+  connection alone, so an image upload or a Create running on it is not cut
+  off. At most one abandoned check per host keeps waiting; until it
+  returns, the next check reports `Timeout` straight away. A link that is
+  really dead is closed by the keepalive within about two keepalive
+  intervals.
+- A check never waits longer than its timeout for the host's connection,
+  even while a reconcile holds it to redial a host that is down.
+
 ## Known gaps
 
 - **No quota or capacity detection.** libvirt reports an out-of-space
   pool or an over-committed host as an ordinary error; the reconciler
   treats it as transient and retries with backoff rather than the slower
-  quota backoff.
+  quota backoff. The health check does report each pool's free space
+  (`poolAvailable`), so a pool that is filling up shows before Creates
+  start failing.
+- **A rotated `kvm-ssh` Secret needs a manager restart.** The key and
+  known_hosts files are read on first use and cached for the life of the
+  process. Replacing them in the Secret therefore changes nothing for
+  later connections, health checks included. After a host key change, the
+  check keeps reporting `SSHHandshakeFailed` until the manager restarts.
 - **Orphan volumes without a domain are invisible to GC.** A crash
   between creating the volumes and defining the domain leaves an overlay
   and seed that `ListByTag` cannot see (it lists domains). They are
   cleaned when the same Proxy reconciles again (Delete removes volumes by
   name even with no domain), but a Proxy CR that is gone for good leaves
   them for `virsh vol-list` to find.
+- **A host removed from the config takes its VMs with it.** `Delete` for a
+  providerID naming an unconfigured host is a no-op success, so the Proxy
+  finishes deleting instead of pinning its finalizer, but the domain and its
+  volumes stay on the hypervisor and orphan GC cannot see them either (it
+  only sweeps configured hosts). Delete the Proxies before retiring a host,
+  or clean up afterwards with `virsh undefine --nvram` plus `virsh vol-delete`.
+  `Get` still fails permanently for such a providerID on purpose: a typo in
+  `hosts` must not look like a missing VM and trigger a rebuild elsewhere.
 - **The first image fetch is synchronous.** It holds a reconcile worker
   for the download (LAN speed, a few hundred MB, once per image per
   host). Other Proxies keep reconciling on the remaining workers.
@@ -254,7 +345,12 @@ The provider logs through controller-runtime's logger, named `kvm`, with
   under *The image cache*.
 - **IP discovery needs either guest traffic or the agent.** A VM on a
   bridge with no DHCP never gets an address and stays `Provisioning`
-  forever; check `virsh console`.
+  forever; check `virsh console`. The same symptom with a booted, addressed
+  VM means the guest agent never came up: check `systemctl status
+  qemu-guest-agent`, `cloud-init status --long` and the vendor script's
+  output in `/var/log/cloud-init-output.log` (a VM without outbound
+  internet cannot install it). A user-data that disables vendor-data
+  (`vendor_data: {enabled: false}`) also removes the agent, by design.
 - **Debian 13 cloud images hang in GRUB under BIOS on a headless VM**
   ([Debian bug #1111240](https://bugs.debian.org/1111240), open, in
   `grub2`): trixie's image ships `terminal_output gfxterm serial` in its
@@ -272,4 +368,6 @@ Unit tests run over an in-memory fake of the libvirt API. The optional
 integration test against a real host is described in
 [testing.md](testing.md#the-kvm-integration-test-real-hypervisor); its
 `-v` output includes the `libvirt call` trace above, at the verbosity
-`KVM_TEST_LOG_LEVEL` selects (default 1).
+`KVM_TEST_LOG_LEVEL` selects (default 1). `TestKVMCheck_realHost` runs the
+health check against the same host and creates nothing:
+`go test -tags=integration -count=1 -v -run TestKVMCheck ./internal/provider/kvm/`.

docs/plans-executions/2026-09-04-1404-kvm-provider.md

@@ -10,7 +10,7 @@ Plan: [2026-09-04-1404-kvm-provider.md](../plans/2026-09-04-1404-kvm-provider.md
 - [x] Step 4 — Provider methods (`kvm.go`) over a fake API
 - [x] Step 5 — Wiring, manifests, samples, docs, CHANGELOG
 - [x] Step 6 — Integration test against a real hypervisor (written and compiled; first real run is part of Step 7)
-- [ ] Step 7 — Manual validation + MR
+- [x] Step 7 — Manual validation + MR
 
 ## Step 0 — Save plan + branch
 
@@ -557,3 +557,117 @@ on the host it needs `virsh undefine --nvram` plus its two volumes by
 hand. The `ListByTag` sweep costs one `virDomainGetMetadata` per
 domain on the host (21 here, ~0.4 ms each over the tunnel), which is
 fine at this scale but is the call to watch on a busy hypervisor.
+
+### Run 7: in-cluster, two defects the integration test could not see
+
+The e2e pass ran the operator inside the kind cluster
+(`egress-proxies-operator:kvm-dev`, built with `make docker-build`,
+loaded with `kind load docker-image`, deployed via the kustomize
+`config/manager` with the `kvm-home` provider pointing at beelink) and
+two Proxies with real cloud-configs: one plain squid, one with a debug
+login user. Both defects come from the difference between the test
+harness and the real thing, not from libvirt.
+
+**Read-only root filesystem.** The manager runs with
+`readOnlyRootFilesystem: true`, and `iso9660.NewWriter` stages the seed
+in a temp directory, so the first Create failed with `creating iso
+writer: read-only file system`. Fix: a 64 MiB `emptyDir` on `/tmp` in
+`config/manager/manager.yaml`. Side finding: the error is classified
+`ErrPermanent` by `writeSeed`, yet `hostPool.call` treated any
+non-libvirt error as a transport failure and dropped the SSH connection,
+so every retry redialled. `call` now leaves the connection alone for
+errors already carrying a provider sentinel (`isClassified`), with a
+test.
+
+**The guest agent never got installed.** Both VMs booted (confirmed on
+the console by hand) but sat in `Provisioning` for 25 minutes: the
+trace showed `virDomainInterfaceAddresses` from the agent source failing
+with `ErrAgentUnresponsive` on every poll and the ARP source empty. The
+integration test had passed with `CloudInit: "#cloud-config\n"`; the
+real Proxies have `packages: [squid]` and `runcmd:`. cloud-init merges
+vendor-data under user-data with its default mergers, which keep the
+user-data's value for any key both define and never append lists, so the
+vendor-data's `packages`/`runcmd` were dropped wholesale. Confirmed
+locally with the dev box's cloud-init 26.1 before changing anything:
+
+```sh
+python3 - <<'EOF'
+from cloudinit import mergers
+user = {"packages": ["squid"], "runcmd": ["systemctl restart squid"]}
+vendor = {"packages": ["qemu-guest-agent"], "runcmd": [["systemctl", "enable", "--now", "qemu-guest-agent"]]}
+print(mergers.construct(mergers.default_mergers()).merge(dict(user), dict(vendor)))
+EOF
+# {'packages': ['squid'], 'runcmd': ['systemctl restart squid']}
+```
+
+Fix: `withGuestAgent` in `seed.go` merges the bootstrap into the Proxy's
+cloud-config itself (append to `packages` and `runcmd`, skip if present,
+re-render with `sigs.k8s.io/yaml`), so vendor-data is no longer defaulted
+and `vendorData` is plain extra data. User-data that is not a cloud-config
+document (script, `#include`, MIME) still gets the bootstrap as
+vendor-data since there is nothing to collide with; a cloud-config that
+does not parse as a mapping fails the Proxy as `ErrPermanent`. The
+rendered document for the squid Proxy passed `cloud-init schema
+--config-file`. The logging test's fixture (`#cloud-config` + a bare
+sentinel line) is not a mapping and had to become `hostname: <sentinel>`.
+
+Rebuilt, reloaded, restarted, deleted and recreated both Proxies: both
+`Ready`/`HEALTHY True` in 2 min 1 s from `kubectl apply`, IP via the
+agent 110 s after `virDomainCreate` (the guest installs squid and the
+agent from the mirrors first), and `curl -x http://<ip>:3128
+https://www.gstatic.com/generate_204` from the dev box returns 204
+through both.
+
+Worth noting: the `#cloud-config` round-trip drops comments and reorders
+keys alphabetically; semantics are preserved (numbers, booleans and
+block scalars survive, checked in tests) but a user diffing the seed
+against their CR will see a different document. Vendor-data supplied via
+`vendorData` is subject to the same cloud-init merge rule and is now
+documented as such. The kind pod also restarted twice during the session
+from "Pod sandbox changed" (the node briefly lost the API server), which
+is unrelated to the operator. Still uncommitted on purpose: the
+`kustomization.yaml` image override (`egress-proxies-operator:kvm-dev`)
+and the `beelink` host entries in `config/manager/providers_config.yaml`
+and `config/samples/providers-config.yaml`, which are this machine's dev
+setup rather than the repo's defaults.
+
+### Run 8: review fixes
+
+A `/code-review high` pass over the branch found five defects, fixed in
+[plan 2026-09-10-2222](../plans/2026-09-10-2222-kvm-review-fixes.md) with
+its own execution log. Two of them land on Run 7's conclusions above and
+supersede them:
+
+- The guest-agent **merge described in Run 7 is gone**. It could not be made
+  safe for user-data that is not a plain cloud-config, and its YAML
+  round-trip re-typed unquoted scalars. The bootstrap now ships as a
+  vendor-data shell script, which cloud-init never merges, and the Proxy's
+  cloud-init reaches the seed byte for byte.
+- The read-only-filesystem fix from Run 7 still stands, but the eager SSH
+  key load it sat next to does not: a missing `kvm-ssh` Secret aborted the
+  whole provider set and crash-looped the manager. The material is read on
+  first use now.
+
+Re-validated the same way: integration test against beelink, both Proxies
+recreated in kind, 204 through squid.
+
+### MR opened
+
+[PR #5](https://gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/pulls/5),
+`feat/kvm-provider` → `main`, 19 commits. `tea` is not installed on the
+dev box and the HTTPS remote has no credential helper, so the push used a
+one-off helper reading a Gitea access token from `~/.config/gitea-token`,
+and the MR was created with the API instead of `tea pr create`:
+
+```sh
+git -c credential.helper='!f() { echo username=kacerr; echo "password=$(tr -d "[:space:]" < ~/.config/gitea-token)"; }; f' \
+  push -u origin feat/kvm-provider
+curl -s -H "Authorization: token $(tr -d '[:space:]' < ~/.config/gitea-token)" \
+  -H 'Content-Type: application/json' -d @pr.json \
+  https://gitea.home.hrajfrisbee.cz/api/v1/repos/kacerr/egress-proxies-operator/pulls
+# pr.json: {"title": ..., "head": "feat/kvm-provider", "base": "main", "body": ...}
+```
+
+The push printed a libsecret "store failed" warning from the system
+credential store; harmless, the token was not persisted anywhere but the
+file.

docs/plans-executions/2026-09-10-2222-kvm-review-fixes.md

@@ -0,0 +1,229 @@
+# Execution: fix the five code-review findings on `feat/kvm-provider`
+Plan: [docs/plans/2026-09-10-2222-kvm-review-fixes.md](../plans/2026-09-10-2222-kvm-review-fixes.md)
+
+- [x] Step 0 — Save plan into the repo
+- [x] Step 1 — Unwedge deletion (finding 1)
+- [x] Step 2 — Validate names that become identifiers (finding 2)
+- [x] Step 3 — Read the SSH material lazily (finding 3)
+- [x] Step 4 — Guest agent via vendor-data script (findings 4 and 5)
+- [x] Step 5 — Validate, record, push
+
+## Step 0 — plan committed
+
+Copied out of `~/.claude/plans/` and committed alone before touching code, as
+the repo's plan rules require.
+
+## Step 1 — Unwedge deletion
+
+Three changes, in the order the bug unfolds.
+
+`reconcileDelete` calls `Get` first and only drops the finalizer when it
+answers not-found, so any permanent error there retried forever. It now
+switches on the class: a permanent `Get` still calls `Delete`, which is
+idempotent, and releases the object when that succeeds. Polling a `Get`
+that cannot converge was the wedge; `Delete` is the only call that can
+still make progress.
+
+`crawl.example.com/skip-cleanup: "true"` abandons the instance and removes
+the finalizer. It is checked **before** the provider lookup, which also
+frees a Proxy whose provider was deleted from the config — a second wedge
+sitting three lines above the reported one, with the same shape. No
+`EventRecorder` exists on the reconciler and none was added, so the
+abandonment is an Info log; `docs/architecture.md` already records that
+decision for the whole controller.
+
+In the KVM provider, `Delete` for a providerID naming an unconfigured host
+now logs and returns success. The precedent is two functions further down:
+`removeVolumes` treats a missing pool the same way, with a comment saying
+outright that refusing "would pin the Proxy's finalizer".
+
+**`Get` deliberately still fails permanently for that case.** Making it
+answer not-found would have been a one-line fix for the wedge, but the live
+reconcile path reads not-found as "the VM is gone" and rebuilds it
+elsewhere. A typo in `hosts` would then quietly duplicate every Proxy on
+that host, with the originals still running where GC cannot see them, since
+`ListByTag` only sweeps configured hosts. The asymmetry is now pinned by a
+test that asserts both halves at once.
+
+## Step 2 — Validate names that become identifiers
+
+One regexp, `^[A-Za-z0-9][A-Za-z0-9_.-]*$`, applied to `hosts[].name` and to
+the `images` map keys. Dots stay legal because a host name like
+`nas1.home.lan` is ordinary and harmless; the character that actually
+breaks things is `/`.
+
+The review only reported the host name. The image key has the same defect
+and is arguably worse: it is not merely embedded in an ID but pasted into a
+file name, `base-<key>.<format>`, so `../etc/x` would be a path traversal
+into the storage pool rather than a parse failure. Both are validated in the
+same pass. Machine-type keys are left alone deliberately — they are only
+ever compared against `placement.machineType`, never embedded anywhere.
+
+Worth noting: this is validation of the operator's own config file, which
+only a cluster admin can write, so the traversal was never a privilege
+boundary. It is still cheaper to reject at load than to explain a stuck
+Proxy later. Both new table cases assert the message text, so the wording
+that names the consequence is now pinned by tests.
+
+## Step 3 — Read the SSH material lazily
+
+`kvm.New` no longer loads the key; `lazyAuth` does, on the first connection,
+caching the result once it works. `registry.Build` is all-or-nothing by
+design, so the old constructor error did not merely fail the KVM provider,
+it exited the manager before any controller started.
+
+The plan said "lazy dialer", and the first attempt was exactly that — one
+closure wrapping both the loading and the dial. It could not be tested: the
+fixture's SSH server is real, but the socket behind it is an echo server, so
+every successful dial dies in the libvirt handshake and the assertion could
+not tell "the key worked" from "the key did not". Splitting the seam in two,
+`lazyAuth` for the material and `lazyDialer` for the wiring, made both
+testable: the healing path against `lazyAuth`, and a separate test that the
+dialer really authenticates, asserted by the error being past SSH and by the
+server having seen the forward request.
+
+Error classes carry the intent. A missing file stays unclassified, so
+`provider.Class` defaults it to transient and a Proxy heals on its own when
+the Secret appears — the common case, since the Secret is usually created
+after the manager. A key that does not parse wraps `ErrPermanent`, because
+retrying cannot fix it. `newSSHAuth` distinguishes the two with
+`errors.Is(err, fs.ErrNotExist)`.
+
+Worth noting: the startup check is kept, just non-fatally, logged with the
+`WARNING:` prefix the repo already uses where logr has no warn level. That
+preserves the diagnostic the eager version was written for. The caching test
+deliberately deletes the key file after a successful load and asserts the
+cached value is reused, which pins the rotation behaviour the docs promise:
+a rotated Secret needs a manager restart. `TestNew_badKeyPathFailsFast` was
+inverted rather than deleted, so the old contract stays visible in the diff
+next to the new one.
+
+## Step 4 — Guest agent via vendor-data script
+
+The review filed two findings against `withGuestAgent` — detection too
+narrow, and YAML re-typing on the round-trip. Reading cloud-init 26.1's
+sources while planning turned both into one finding: the operator should
+not be editing the user's document at all.
+
+Vendor-data gets its own handler chain (`stages.py:_default_vendordata_handlers`),
+with separate script and cloud-config paths. A vendor-data **script** is
+therefore written to the vendor script directory and run by
+`cc_scripts_vendor`, never touching the config merger that drops colliding
+keys. On the final-stage module list it lands after
+`package_update_upgrade_install` and before the user's `runcmd`, so the
+agent now comes up *earlier* than the old `runcmd` append did.
+
+That also closed the door the previous approach had left ajar. The merge
+cannot be rescued from inside either document: `merge_how` in user-data
+governs only the first source, and `merge_how` in vendor-data is popped by
+`dict_extract_mergers` and discarded before the document is written. There
+was no fix that kept the merge.
+
+`withGuestAgent`, `isCloudConfig` and `appendUnique` are gone, along with
+seed.go's `sigs.k8s.io/yaml` dependency. `installGuestAgent` (default true)
+turns the script off. When `vendorData` is also configured the two ship as
+a MIME multipart, with the user's document declared `text/plain` — a type
+cloud-init re-sniffs from the content's own header line, so this code never
+classifies what the user wrote.
+
+Before deleting the old path, its damage was captured rather than taken on
+trust. Round-tripping a plausible cloud-config through it:
+
+```sh
+# temporary test in the package, deleted afterwards
+go test -run TestZZProbeRetyping -v ./internal/provider/kvm/
+# in:  version: 1.10
+# out: version: 1.1
+```
+
+`1.10` arriving as `1.1` is worse than the reviewer's `passwd` example,
+since nothing about it looks wrong in the seed. Comments were dropped and
+keys reordered on the same pass.
+
+The generated vendor-data was then checked against the cloud-init actually
+installed on this machine, rather than against the docs:
+
+```sh
+python3 -c 'from cloudinit import user_data, handlers; ...'
+# vendor-script:    multipart=False  sniffed=text/x-shellscript
+# vendor-multipart: multipart=True   part 1 text/plain -> text/cloud-config
+#                                    part 2 text/x-shellscript
+```
+
+and the script itself through `sh -n`.
+
+Worth noting: the script has to choose a package manager where cloud-init's
+`packages:` key was distro-agnostic. It covers apt, dnf, yum, zypper and
+apk, and both systemd and OpenRC for the enable step, but an image outside
+that set now needs `installGuestAgent: false` and the agent baked in. The
+seed's user-data is byte-identical to the Proxy's `cloudInit` from here on,
+which is what the new `TestBuildSeedISO_userDataIsNeverRewritten` pins
+across seven formats. One consequence worth remembering: a user-data that
+sets `vendor_data: {enabled: false}` disables the agent, and cloud-init
+documents that as intended behaviour.
+
+## Step 5 — Validation
+
+Unit suite, `go vet` and `gofmt` clean across the repo. Then the two real
+runs.
+
+**Integration against beelink**, unchanged environment from the KVM plan's
+log, passed in 22 s on a warm cache, with
+`virDomainInterfaceAddresses source=DomainInterfaceAddressesSrcAgent
+interfaces=2` — the agent installed by the new vendor script answering, on
+a stock Debian 13 image.
+
+**In-cluster**, both Proxies deleted and recreated against the rebuilt
+image: Ready with an address 2 min 3 s after `kubectl apply`, 204 through
+squid from the dev box for both. Inside the debug VM:
+
+```sh
+ssh novakj@192.168.0.94 'systemctl is-active qemu-guest-agent; cloud-init status; \
+  sudo head -4 /var/lib/cloud/instance/user-data.txt; sudo ls /var/lib/cloud/instance/scripts/vendor/'
+# active
+# status: done
+# #cloud-config
+# # Login for debugging: user "novakj", password "kvm-dev", passwordless sudo.
+# ...
+# part-001
+```
+
+Two things worth having seen with your own eyes there: the user's **comment
+lines survived** into the guest, which the old merge stripped, and the
+script really did land in the vendor directory rather than being merged
+into the user's config.
+
+**Negative check for the Secret**, done by pointing the deployment's volume
+at a name that does not exist rather than deleting the real Secret:
+
+```sh
+kubectl -n egress-proxies-operator-system patch deploy egress-proxies-operator-controller-manager \
+  --type=json -p '[{"op":"replace","path":"/spec/template/spec/volumes/1/secret/secretName","value":"kvm-ssh-absent"}]'
+```
+
+The manager came up and stayed up, 0 restarts, logging
+`WARNING: kvm ssh material is not usable ... open /etc/proxy-operator/kvm/id_ed25519:
+no such file or directory`. The Kubernetes Proxies kept reconciling; the
+KVM ones failed per reconcile and retried. Before this branch the same
+config exited the process before any controller started. Patched back
+afterwards, everything Ready.
+
+Worth noting: the KVM Proxies kept reporting `Ready` during that window
+rather than degrading visibly. That is existing behaviour, not a
+regression — a transient provider error rides the workqueue and does not
+clear a healthy status — but it does mean the only signal of a missing
+Secret is the log and the reconcile errors. Also, the ListByTag sweep on
+beelink still reports `managed=2` where one domain is an orphan from an
+older failed run; unrelated to this work, and the same leftover noted in
+the KVM plan's Run 6.
+
+Worth noting: the review filed this as a KVM finding, but the fix is mostly
+in the controller, because KVM is the only provider that resolves a
+providerID segment against operator config at all. GCP and Kubernetes hand
+the lookup to the backend, which answers 404 for a retired zone or a deleted
+namespace, so they reach not-found and release on their own. Anything else
+that classifies permanently on the delete path — revoked RBAC on the
+Kubernetes provider is the realistic one — was equally stuck before this
+step and is what the annotation is really for. The KVM table case
+`"host not configured"` moved out of `TestDelete_errors` into its own named
+test, since it is no longer an error.

docs/plans-executions/2026-09-11-1526-provider-health-checks.md

@@ -0,0 +1,378 @@
+# Execution: provider health checks
+Plan: [docs/plans/2026-09-11-1526-provider-health-checks.md](../plans/2026-09-11-1526-provider-health-checks.md)
+
+- [x] Step 0 — Save the plan
+- [x] Step 1 — ProviderHealth API and CRD
+- [x] Step 2 — Contract plus the kubernetes and GCP checks
+- [x] Step 3 — KVM check
+- [x] Step 4 — Monitor
+- [x] Step 5 — Metrics and wiring
+- [x] Step 6 — Docs
+- [x] Step 7 — Validate, record, push, open the PR
+
+## Step 0 — Save the plan
+
+Branched off `main` (at the PR #5 merge, `eb8f075`). The plan was copied out of
+`~/.claude/plans/` and committed on its own, before any code, as the repo's plan rules
+require:
+
+```bash
+git checkout -b feat/provider-health main
+cp ~/.claude/plans/now-i-would-like-flickering-wall.md docs/plans/2026-09-11-1526-provider-health-checks.md
+```
+
+The uncommitted `.claude/settings.json` change was left unstaged and stays out of this
+branch.
+
+## Step 1 — ProviderHealth API and CRD
+
+`kubebuilder` isn't installed here (only `bin/controller-gen`, `bin/kustomize` and
+`bin/setup-envtest` are), so `api/v1alpha1/providerhealth_types.go` and the `PROJECT`
+entry were written by hand instead of with `kubebuilder create api`. For a cluster-scoped
+resource `PROJECT` simply omits `namespaced`, which is what kubebuilder itself writes.
+
+The new types are registered in the existing `AddKnownTypes` call in `proxy_types.go`,
+not in a second `init()`.
+
+The deepcopy code, the CRD and the RBAC come from the usual generators:
+
+```bash
+make generate manifests
+go build ./... && go vet ./api/... && go test -short ./api/...
+```
+
+To confirm the new CRD and the viewer role actually reach the rendered install, I built
+the kustomization and grepped for them:
+
+```bash
+bin/kustomize build config/default | grep -n providerhealth
+# providerhealths.crawl.example.com (scope: Cluster, shortName ph)
+# egress-proxies-operator-providerhealth-viewer-role
+```
+
+Worth noting:
+- The only role added is the viewer role. The operator is the object's only writer, so
+  the admin and editor roles kubebuilder scaffolds for Proxy were not added.
+- The type file has no unit test, since it has no logic. The schema, including the
+  required `ready`/`lastTransitionTime` and the list-map keys, is proven by the envtest
+  round-trip in Step 4.
+- `config/rbac/role.yaml` is unchanged so far. The manager's own permissions arrive in
+  Step 4 with the Monitor's RBAC markers.
+
+## Step 2 — Contract plus the kubernetes and GCP checks
+
+`provider.Checker` is an optional interface in `internal/provider/check.go`, kept out of
+`Provider`. Each implementation has a compile-time assertion, because the monitor's
+type assertion would otherwise fail silently.
+
+- **kubernetes:** lists at most one managed Pod cluster-wide, using the same client, verb
+  and scope as `ListByTag`. The target is `apiserver`, at the `restCfg.Host` now stored
+  in the new `apiHost` field.
+- **GCP:** calls a new seam method, `Probe`. `realInstances.Probe` lives in `check.go`,
+  next to its only caller. `ListByTag`'s `Sprintf` filter became the shared
+  `managedFilter` constant, so the check and the GC sweep can't drift apart.
+
+Before relying on "MaxResults is only a page size", I checked the SDK's REST
+`AggregatedList` in the module cache:
+
+```bash
+grep -n "func (c \*instancesRESTClient) AggregatedList" -A80 \
+  "$(go list -m -f '{{.Dir}}' cloud.google.com/go/compute)"/apiv1/instances_client.go
+# InternalFetch: req.MaxResults = pageSize ... return elems, resp.GetNextPageToken(), nil
+```
+
+```bash
+go build ./... && go vet ./internal/provider/... && go test -race -count=1 ./internal/provider/...
+```
+
+Worth noting:
+- The planned `TestRealInstances_Probe_fetchesOnePage` became a three-case
+  `TestRealInstances_Probe` table.
+  - It keeps the first page's `nextPageToken`, and asserts exactly one request with
+    `maxResults=1` for each case.
+  - "empty project" covers `iterator.Done`.
+  - "api error keeps its status" proves the real client returns a `*googleapi.Error`, so
+    a disabled Compute API is reported as `Forbidden` rather than `Unreachable`.
+- Once a context has ended, both checks report `Timeout` whatever the error says. The
+  monitor's timeout is the usual cause, and the underlying error only repeats it.
+
+## Step 3 — KVM check
+
+`conn.go` now wraps each of the four dial stages in a `dialError`: ssh-material, tcp,
+ssh-handshake and libvirt-connect. `Error()` returns the inner message unchanged and
+`Unwrap()` returns the cause, so `classify`/`isClassified` and every existing
+message-matching test behave as before. `TestDialError_keepsClassification` pins this
+down.
+
+The three new RPCs went into `libvirtAPI`, `tracingAPI` and `fakeAPI`. The method-count
+guard in `TestTracingAPI_logsEveryMethod` fails until a traced row exists for each new
+method, which is what forced the three new rows.
+
+`Check` runs one goroutine per host, over the shared connection, through
+`hostPool.call`. The pool verdict is decided only after `call` returns, because
+returning it as an error from inside the callback would make `call` drop a perfectly
+good connection.
+
+Build, vet (also with the integration tag, since that file changed) and race tests:
+
+```bash
+go build ./internal/provider/... && go vet ./internal/provider/kvm/ \
+  && go vet -tags=integration ./internal/provider/kvm/ \
+  && go test -race -count=1 ./internal/provider/kvm/
+```
+
+Worth noting:
+- **Pool state wording:** `poolState` uses virsh's words (`running`, `inactive`, …)
+  rather than go-libvirt's constant names, so it reads like `virsh pool-list` output.
+- **Details:** each host also reports its SSH `user`. That is the first thing to check
+  when the reason is `SSHHandshakeFailed`.
+- **Tests renamed or merged:**
+  - The planned `TestSSHAuth_dial_stages` is `TestLazySSH_dial_tagsTheFailingStage`. It
+    dials through `lazySSH`, so the ssh-material stage is covered as well as the three
+    stages the in-process SSH server can trigger.
+  - The pool-inactive and pool-missing cases are merged into `TestCheck_libvirtVerdicts`,
+    which also covers a daemon error. Each case runs two checks and asserts a single
+    dial.
+- **Tests added beyond the plan:**
+  - `TestCheck_timeout`: a daemon that stops answering gives `Timeout`.
+  - `TestDialError_keepsClassification`.
+- **Not run yet:** `TestKVMCheck_realHost` in the integration file only compiles here.
+  It needs beelink, so it goes on the Step 7 checklist for the user to run.
+
+## Step 4 — Monitor
+
+`internal/providerhealth` has the `Monitor` runnable plus the pure `buildStatus`.
+
+- **Scheduling:** the Monitor is leader-only. It cleans up once at start, checks
+  immediately, then checks on a ticker.
+- **Each round:** providers are checked concurrently under one timeout, then logged and
+  published one after another. A failed write is logged, and the next round writes
+  again.
+- **Reads:** objects are read through the reader it is given. `main.go` passes the
+  manager's API reader (Step 5).
+
+The first `make manifests` left `config/rbac/role.yaml` unchanged, with no error.
+I had put the `+kubebuilder:rbac` markers inside `Monitor`'s doc comment, and
+controller-gen only reads RBAC markers from package-level comment groups. So they were
+silently attached to the type and ignored. An empty `git diff config/rbac/role.yaml` was
+the only sign. Moving them into their own comment block fixed it:
+
+```bash
+make manifests && git diff config/rbac/role.yaml
+# + providerhealths: create, delete, get, list
+# + providerhealths/status: get, patch
+```
+
+Vet, race tests and the envtest round-trip (using the `bin/k8s/1.36.2-linux-amd64`
+binaries `make setup-envtest` had already fetched):
+
+```bash
+go vet ./internal/providerhealth/ && go test -race -count=1 -v ./internal/providerhealth/
+# --- PASS: TestMonitor_envtestRoundTrip (6.45s) ... ok
+```
+
+Worth noting:
+- **Fake client scope:** the fake client needs a RESTMapper that registers
+  ProviderHealth as `RESTScopeRoot`. Its default mapper treats every kind as namespaced.
+- **Why the envtest calls `publish` directly:** `checkOnce` only logs write errors, so a
+  schema rejection would look like a missing object. Calling `publish` makes the API
+  server's answer the test failure.
+  - Its first round writes a 2000-character message; the stored message is truncated to
+    1024.
+  - Its second round proves that only the target that flipped gets a new
+    `lastTransitionTime`.
+- **No guard against a hung checker:** the Monitor trusts `Checker`'s contract to return
+  once the context ends, and doesn't abandon a checker that ignores it. All three
+  checkers stop when the context ends: KVM through `awaitOrCancel`, and kubernetes and
+  GCP through their clients.
+- **Test renamed:** the planned `TestMonitor_timeoutBoundsBlockingChecker` is
+  `TestMonitor_checkOnce_timeoutBoundsEachRound`. It also asserts that a fast provider
+  still publishes next to a hung one.
+
+## Step 5 — Metrics and wiring
+
+`metrics.RegisterProviderTargets` registers a scrape-time collector for three gauges,
+each labelled `{provider,target}`: `up`, check duration and check timestamp.
+`Register`'s signature is unchanged.
+
+`cmd/main.go`:
+- gains `--provider-check-interval` (1m) and `--provider-check-timeout` (15s);
+- validates them before even loading the providers config;
+- collects the `Checker`s from the unwrapped providers in config order, just before the
+  metrics and tracing wrap loop;
+- adds the Monitor with the manager's API reader;
+- registers the collector through a small `providerTargetSamples` closure next to
+  `proxyPhaseCounts`.
+
+Build, vet and metrics tests:
+
+```bash
+go build ./... && go vet ./cmd/... ./internal/metrics/ && go test -race -count=1 ./internal/metrics/
+```
+
+To confirm bad flags fail fast, before anything touches the cluster, I ran the binary
+with a 5s interval. zap's development mode prints a stack trace after the error, hence
+the grep:
+
+```bash
+go run ./cmd --providers-config=/dev/null --provider-check-interval=5s 2>&1 | grep -m1 "Invalid provider check flags"
+# ERROR setup Invalid provider check flags {"error": "--provider-check-interval must be 0 (off) or at least 10s, got 5s"}
+go vet -tags=e2e ./test/e2e/
+```
+
+The e2e suite gets a new ordered spec, "should publish provider health". It waits for
+`providerhealth kubernetes` to report `Ready=True`. It runs before the existing metrics
+spec, which now also asserts
+`proxy_operator_provider_target_up{provider="kubernetes",target="apiserver"} 1`. By then
+the first round is known to be done, so the series is certain to exist.
+
+Worth noting:
+- **Main-only wiring:** `providerTargetSamples` is an 8-line loop in `package main`,
+  which has no test file. The e2e metrics assertion is its only coverage.
+- **The KVM entry isn't asserted in e2e:** the default providers config also ships
+  `kvm-home` (beelink), which would report `SSHMaterialUnavailable` in kind. That entry
+  is the user's own host, and asserting it would break the suite as soon as the default
+  config changes.
+- **Manifest:** `config/manager/manager.yaml` only gained a comment about the two flags.
+  The defaults need no args.
+
+## Step 6 — Docs
+
+- **`docs/architecture.md`:**
+  - component rows for the CRD, the `Checker` contract and the monitor;
+  - a new §11 covering the provider table, the loop diagram, the write, read and
+    decorator choices, the gauges, and sample `kubectl get ph` output;
+  - cross-references from the metrics (§9) and tracing (§10) sections;
+  - four Decisions entries.
+- **`docs/kvm-provider.md`:**
+  - a "Health check" section with the reason → meaning → where-to-look table and the
+    two consequences of sharing the connection;
+  - under Known gaps, the capacity bullet now points at `poolAvailable`, and a new
+    bullet says a rotated `kvm-ssh` Secret needs a restart;
+  - the Testing section names `TestKVMCheck_realHost`.
+- **`docs/testing.md`:**
+  - suite rows for the three `check_test.go` files, `internal/providerhealth` (unit and
+    envtest) and the metrics gauges;
+  - "17 libvirt calls" is now 20;
+  - the integration and e2e sections describe the new tests.
+- **`CLAUDE.md`:**
+  - the Architecture TODO now points at `docs/architecture.md`;
+  - the stale "only `go.mod` exists" gotcha is replaced by a note that the "What this
+    is" and "Commands" sections predate the code, plus Step 4's RBAC-marker gotcha.
+
+Worth noting:
+- `docs/api.md` is untouched. It documents the discovery HTTP API, which this change
+  doesn't alter.
+- `CLAUDE.md`'s "What this is" and "Commands" sections were left stale on purpose.
+  Rewriting them is outside this plan, so the new gotcha flags them instead of leaving
+  them silently wrong.
+
+## Step 7 — Validate, record, push, open the PR
+
+The full suite runs codegen, `go fmt` and `go vet`, then `go test -race` on every
+package, with both envtest suites included:
+
+```bash
+make test
+# every package ok; internal/providerhealth 92.5 %, internal/provider/kvm 92.0 % coverage
+```
+
+`go fmt` rewrote nothing, and `git status` showed only the Step 6 docs.
+
+`make lint` could not run, for a reason unrelated to this branch. The Makefile rebuilt
+the custom golangci-lint, which reports "built with go1.27.0", and its bundled
+staticcheck (honnef.co/go/tools v0.7.0) panics on the standard library before it
+reaches any repo package:
+
+```bash
+make lint
+# buildir: package "poll" (isInitialPkg: false, needAnalyzeSource: true): unexpected expr: *ast.KeyValueExpr
+# make: *** [Makefile:107: lint] Error 3
+```
+
+The local toolchain is go1.27.0 while `go.mod` says 1.26.0; the otel-tracing plan's
+log shows `make lint` running normally before the toolchain moved. Retrying with
+`GOTOOLCHAIN=go1.26.0` would have meant downloading a toolchain, and the user declined
+that. `go vet` is clean in all three configurations: plain inside `make test`, plus
+`-tags=integration` and `-tags=e2e`.
+
+The kind e2e suite creates and deletes its own `egress-proxies-operator-test-e2e`
+cluster. I checked `KIND_CLUSTER` first, because the two clusters already on the
+machine (`kind`, `proxy-operator-demo`) must not be torn down by `cleanup-test-e2e`:
+
+```bash
+make test-e2e > <scratch>/e2e.log 2>&1
+# Ran 3 of 5 Specs in 86.210 seconds
+# SUCCESS! -- 3 Passed | 0 Failed | 0 Pending | 2 Skipped
+```
+
+In the deployed manager, the `kubernetes` ProviderHealth reported `Ready=True` about a
+second after the pod was up (20:30:05.6 → 20:30:06.7). The metrics scrape that followed
+contained `proxy_operator_provider_target_up{provider="kubernetes",target="apiserver"} 1`.
+That exercises the generated RBAC, the API-reader path and the `main.go` gauge wiring
+end to end. The two skipped specs are the Tempo-gated tracing specs; `TEMPO_URL` was not
+set.
+
+### Follow-up — the check must not cost the reconciles their connection
+
+Reviewing the shared-connection decision after the automated validation turned up two
+ways the health check could hurt the work it shares the link with, both of which the
+plan's "a check against a hung host drops the shared connection" line accepted too
+easily:
+
+- A check that timed out went through `hostPool.call`, whose cancel path drops the
+  connection. A `libvirtd` slow to answer one `ConnectGetLibVersion` would therefore
+  cut off an image upload or a `Create` running on the same link.
+- `hostEntry` was guarded by a `sync.Mutex`, so waiting for a host's slot could not be
+  interrupted. A reconcile redialing a host that is down holds the slot for as long as
+  the SSH dial takes — minutes on a blackholed address — and the whole check round
+  would wait behind it, past its own timeout.
+
+So the check now calls `hostPool.probe` instead of `call`: on `ctx` ending it abandons
+the RPC and leaves the connection alone (a link that is really dead is closed by the
+keepalive anyway, which ends the abandoned `fn` too). One abandoned probe per host is
+allowed to stay waiting — `hostEntry.probing` is an `atomic.Bool`, not a counter, so a
+daemon that never answers cannot pile up goroutines; further probes get the new
+`errProbeBusy`, which `checkReason` maps to `Timeout`. `hostEntry.mu` became a one-slot
+channel with an `acquire(ctx)` that only bounds *waiting*: an uncontended caller takes a
+free slot even with an expired context, so `call`'s behaviour is unchanged where nothing
+is contended.
+
+The second change came out of the same pass: `--provider-check-timeout` defaulted to a
+flat 15s while the flag validation rejected a timeout longer than the interval, so
+`--provider-check-interval=12s` alone failed to start. The flag now defaults to 0 and
+`Monitor.setDefaults` derives `min(15s, Interval)`; `ValidateSchedule` accepts 0 and
+rejects only negatives and timeouts past the interval.
+
+Re-verified with the same commands as above:
+
+```bash
+go vet ./...
+go test -race -count=1 ./internal/provider/kvm/ ./internal/providerhealth/
+make test
+# every package ok; internal/provider/kvm 92.2 %, internal/providerhealth 93.7 % coverage
+```
+
+Worth noting:
+- `make lint` still cannot run, for the toolchain reason recorded above; unchanged by
+  this follow-up.
+- The two new `kvm` tests are timing-shaped by necessity — they assert that a caller
+  with a 50ms deadline does not wait out a dial that is blocked on a channel. The
+  assertions are one-sided (`> 1s` fails) so a slow machine does not make them flaky.
+- `docs/architecture.md`, `docs/kvm-provider.md` and `docs/testing.md` were corrected in
+  the same commit: the old text promised the drop-on-timeout behaviour that is now gone.
+- The real-hypervisor commands from the plan's Verification section are still the user's
+  to run; nothing here replaces them.
+
+### Real-hypervisor validation and PR
+
+The user ran the plan's real-hypervisor checks against beelink (the `TestKVMCheck`
+integration test plus the manual `ProviderHealth` and metrics checks) and confirmed
+they passed. After that the CHANGELOG entry went in, and the branch was pushed with
+the token-file credential helper. The PR against `main` was opened through the Gitea
+API because `tea` isn't installed.
+
+Worth noting:
+- The branch also carries `docs/plans/2026-09-16-0841-provider-blueprint.md`, on
+  purpose: that plan reaches `main` with this merge, and the blueprint work starts on a
+  new branch afterwards.

docs/plans/2026-09-10-2222-kvm-review-fixes.md

@@ -0,0 +1,195 @@
+# Plan: Fix the five code-review findings on `feat/kvm-provider`
+**Created:** 2026-09-10 22:22 UTC
+
+## Context
+
+PR #5 (`feat/kvm-provider` → `main`, 19 commits) is open and awaiting merge. A
+`/code-review high` pass over the branch returned five confirmed findings: two that can
+pin a Proxy's finalizer forever, one startup crash-loop, and two in the guest-agent merge
+committed earlier today. All five are fixed on the same branch so the PR updates in place,
+then re-validated against the real hypervisor before merging.
+
+Four decisions, made by the user during planning:
+
+1. **Wedged deletes** — attempt `Delete` even when `Get` fails permanently, make an
+   unconfigured KVM host a no-op delete, and add an annotation that abandons cleanup for
+   whatever still cannot finish.
+2. **Missing `kvm-ssh` Secret** — read the SSH material lazily, so only KVM Proxies fail.
+3. **Guest agent** — stop rewriting the user's cloud-config. Ship the bootstrap as a
+   vendor-data shell script instead.
+4. Validation is unit tests, the integration test against beelink, and the kind e2e. The
+   uncommitted dev-only files stay exactly as they are and out of the PR.
+
+## Why the guest-agent fix is a rewrite, not a patch
+
+Both cloud-init findings come from the same root cause: this operator edits the user's
+document. Reading cloud-init 26.1's sources settles that it never has to.
+
+Vendor-data is consumed by its **own** handler chain with separate script and config paths
+(`stages.py:_default_vendordata_handlers`). A vendor-data **shell script** is therefore
+written to the vendor script directory and run by `cc_scripts_vendor`, never passing
+through the config merger that drops colliding keys. On the final-stage module list it
+runs after `package_update_upgrade_install` and before the user's own `runcmd`, so the
+agent comes up early regardless of what the user's config does.
+
+That also closes the door the previous approach left open. The merge cannot be rescued
+from inside the documents: `merge_how` in the user-data governs only the first source, and
+`merge_how` in vendor-data is popped and discarded by `dict_extract_mergers` before the
+document is ever written. There is no supported way to make vendor-data cloud-config keys
+survive a user-data that defines the same key.
+
+Consequences: user-data reaches the seed byte for byte in any format, so Jinja templates,
+MIME archives, gzip payloads and shell scripts stop being special cases (finding 4), and
+nothing re-serialises YAML, so unquoted scalars keep their type (finding 5). The cost is
+that the script picks the package manager itself where cloud-init's `packages` key was
+distro-agnostic.
+
+## Why the delete fix is shaped this way
+
+KVM is the only provider that resolves a providerID segment against operator config. GCP
+and Kubernetes let the backend answer "not found", so retiring a location yields
+`ErrNotFound` and the finalizer drops. `Provider.host()`
+([kvm.go:118-139](internal/provider/kvm/kvm.go#L118-L139)) returns `ErrPermanent`, and
+`reconcileDelete` ([proxy_controller.go:281-291](internal/controller/proxy_controller.go#L281-L291))
+drops the finalizer only on `ErrNotFound`. `deletionFailure`
+([:327-336](internal/controller/proxy_controller.go#L327-L336)) says in its own comment
+that it retries "until an operator intervenes", and no intervention mechanism exists. The
+annotation is that mechanism, and it also rescues the unconfigured-provider wedge at
+[:276-279](internal/controller/proxy_controller.go#L276-L279).
+
+`Get` keeps returning `ErrPermanent` for an unknown host deliberately. Turning it into
+`ErrNotFound` would make a typo in the host list look like a missing VM on the live path,
+and the reconciler would recreate every affected Proxy elsewhere while the originals keep
+running where garbage collection cannot see them.
+
+## Steps
+
+One commit per step on `feat/kvm-provider`, each ending with
+`Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>`.
+
+### Step 0 — Save this plan into the repo
+Copy to `docs/plans/2026-09-10-2222-kvm-review-fixes.md` with the same header timestamp and
+commit it alone, before any code change. Start
+`docs/plans-executions/2026-09-10-2222-kvm-review-fixes.md` with the status checklist when
+Step 1 lands.
+
+### Step 1 — Unwedge deletion (finding 1)
+- `api/v1alpha1/proxy_types.go`: add `AnnotationSkipCleanup = "crawl.example.com/skip-cleanup"`
+  beside `AnnotationSpecHash` ([:59-63](api/v1alpha1/proxy_types.go#L59-L63)), documented as
+  abandoning the backend instance.
+- `internal/controller/proxy_controller.go`, `reconcileDelete`: honour the annotation after
+  the empty-providerID branch and **before** the provider lookup, so it also covers a
+  provider that is no longer configured. Log the abandoned providerID at Info.
+  `ProxyReconciler` has no `EventRecorder`; do not add one for this.
+- Same function: replace the two `if`s after `Get` with a switch. On
+  `provider.Class(err) == provider.ErrPermanent`, call `Delete` anyway, since it is
+  idempotent, and remove the finalizer when it succeeds rather than requeueing into a `Get`
+  that can never converge. On failure, `deletionFailure(errors.Join(getErr, delErr))`.
+  Guard on `err != nil`, since `Class(nil)` is nil.
+- `internal/provider/kvm/kvm.go`, `Delete`: an unconfigured host segment means there is
+  nothing this operator can clean, so log it and return success, mirroring `removeVolumes`
+  ([kvm.go:506-514](internal/provider/kvm/kvm.go#L506-L514)) for a missing pool. `Get` is
+  unchanged.
+- Tests: extend the deletion table in `internal/controller/reconcile_test.go`
+  ([:357-393](internal/controller/reconcile_test.go#L357-L393)) with permanent-Get plus
+  succeeding Delete, permanent-Get plus failing Delete, the annotation with the provider
+  never called, and the annotation on an unconfigured provider. In `kvm_test.go`, the
+  `"host not configured"` case at [:737](internal/provider/kvm/kvm_test.go#L737) flips from
+  `ErrPermanent` to success; add a named test beside `TestDelete_missingPoolIsNotAnError`
+  ([:715](internal/provider/kvm/kvm_test.go#L715)).
+- Docs: the annotation in `README.md` (~line 163) and `docs/architecture.md` (~line 487);
+  a *Known gaps* entry in `docs/kvm-provider.md` saying a domain on a host removed from
+  config can no longer be cleaned up by the operator, the Proxy still finishes deleting,
+  and the VM must be removed by hand because garbage collection only sweeps configured
+  hosts.
+
+### Step 2 — Validate names that become identifiers (finding 2)
+- `internal/provider/config_kvm.go`: add a name regexp beside `sha256Hex`
+  ([:6](internal/provider/config_kvm.go#L6)) allowing letters, digits, dot, dash and
+  underscore. Apply it to `hosts[].name` after the existing empty check
+  ([:129-131](internal/provider/config_kvm.go#L129-L131)) and to each `images` map key,
+  which becomes the cached volume name `base-<key>.<format>`
+  ([image.go:49-55](internal/provider/kvm/image.go#L49-L55)) and half of the image-cache
+  lock key ([image.go:96](internal/provider/kvm/image.go#L96)). `machineTypes` keys are only
+  compared, never embedded, so leave them.
+- Keep the file's first-error-wins style and `fmt.Sprintf` field paths. The host message
+  should name the consequence: the value becomes part of `hosts/<name>/domains/<vm>`, which
+  `parseProviderID` splits on `/` ([id.go:11-21](internal/provider/kvm/id.go#L11-L21)).
+- Tests: two entries in the `TestLoadConfig_kvmInvalid` table
+  ([config_kvm_test.go:76](internal/provider/config_kvm_test.go#L76)), existing
+  name/yaml/wantErrSub shape.
+
+### Step 3 — Read the SSH material lazily (finding 3)
+- `internal/provider/kvm/conn.go`: keep `newSSHAuth`, and add a lazy dialer that builds it
+  on first dial and caches it once it succeeds, so a `kvm-ssh` Secret created after the
+  manager starts heals on the next reconcile with no restart.
+- `internal/provider/kvm/kvm.go`, `New`: stop calling `newSSHAuth`; pass the lazy dialer.
+  Use the context parameter it currently ignores to log a warning when either file is
+  unreadable at startup, keeping the diagnostic without the exit.
+- Error classes: a missing or unreadable file stays unclassified, so `Class` defaults it to
+  transient and the Proxy retries until the Secret appears; an unparseable key or malformed
+  known_hosts wraps `ErrPermanent`. Distinguish with `errors.Is(err, fs.ErrNotExist)`.
+- Fix the two doc comments promising startup failure
+  ([kvm.go:79-81](internal/provider/kvm/kvm.go#L79-L81),
+  [conn.go:182-185](internal/provider/kvm/conn.go#L182-L185)) and the matching claim in
+  `docs/kvm-provider.md` (~lines 107-109). `DISCOVERY_TOKEN` already sets the house
+  convention for an optional mount: degrade with a loud warning.
+- Tests in `conn_test.go`, which already writes key files: `New` succeeds with no key
+  present, the first dial fails transiently, a dial after the file appears succeeds, and a
+  corrupt key is permanent.
+
+### Step 4 — Guest agent via vendor-data script (findings 4 and 5)
+- `internal/provider/kvm/seed.go`: delete `withGuestAgent`, `isCloudConfig`, `appendUnique`
+  and the `sigs.k8s.io/yaml` import. User-data goes onto the seed exactly as the Proxy
+  supplied it.
+- Add the bootstrap as a `#!/bin/sh` vendor-data script: skip when the agent is already
+  present, otherwise install `qemu-guest-agent` through whichever of `apt-get`, `dnf`,
+  `yum`, `zypper` or `apk` exists, then `systemctl enable --now`. Run `apt-get update`
+  before installing, since nothing else guarantees a fresh index when the user's config has
+  no `packages` key. Keep it idempotent and let it log to stdout, which lands in
+  `cloud-init-output.log`.
+- When the provider config also sets `vendorData`, wrap both into a MIME multipart
+  vendor-data: the user's document as one part, the script as another. Cloud-init detects
+  multipart by finding `mime-version:` in the first 4096 bytes, so the header must lead.
+- `internal/provider/config_kvm.go`: add `installGuestAgent`, defaulting to on. Off means
+  the seed carries only the user's own vendor-data and IP discovery falls back to the host
+  ARP table. Nothing fails permanently any more, so the error path in `writeSeed`
+  ([kvm.go:252-255](internal/provider/kvm/kvm.go#L252-L255)) narrows to MIME assembly.
+- Tests in `seed_test.go`: user-data byte-identical for cloud-config, Jinja, MIME, gzip and
+  script inputs; the script present in vendor-data; the opt-out; the multipart case
+  including the `mime-version:` position; keep the existing ISO tests.
+- Docs: rewrite the *What lands on the host* passage added earlier today, the `vendorData`
+  row and a new `installGuestAgent` row in the config table, and the ARP note in
+  *Known gaps*. Cite the cloud-init behaviour that forced the change.
+
+### Step 5 — Validate, record, push
+- `gofmt -l internal/`, `go vet ./...`, `go test -short ./...`.
+- `make test-integration-kvm` against beelink, environment as recorded in the KVM execution
+  log.
+- Rebuild the image, `kind load`, restart the manager, delete and recreate both Proxies,
+  curl a test URL through each.
+- `CHANGELOG.md` entry at the top with a real `date` timestamp, noting it supersedes the
+  merge approach described in today's earlier entry. Add a Run 8 section to
+  `docs/plans-executions/2026-09-04-1404-kvm-provider.md` plus this plan's own execution
+  file. Push with the token helper; PR #5 updates in place.
+
+## Verification
+- Unit: the controller cases prove a permanent `Get` no longer pins the finalizer and that
+  the annotation short-circuits before the provider is touched. `TestDelete_*` proves an
+  unconfigured host is a clean delete while `TestGet_*` stays permanent, which is what keeps
+  the live path from recreating VMs.
+- Seed: assert the uploaded user-data equals the Proxy's `cloudInit` byte for byte, which is
+  the property the old merge could not offer.
+- Integration on beelink: Create, IP via the guest agent, ListByTag, Delete. In the guest,
+  `cloud-init status --long` clean and `systemctl is-active qemu-guest-agent` active.
+- In-cluster: both Proxies reach Ready with an IP and return 204 through squid. Their
+  cloud-config defines its own `packages`, which is the exact case that failed in Run 7.
+- Negative check in kind: rename the `kvm-ssh` Secret away and restart. The pod must stay
+  up, the Kubernetes Proxies must stay Ready, and only the KVM Proxies degrade. Restore the
+  Secret and confirm they recover without a restart.
+
+## Out of scope
+- The two items the review noted but did not report: deleting a cached base image while
+  running VMs back onto it, and an ad-hoc image URL always being treated as qcow2. Both are
+  documented known gaps and neither is a regression from this branch.
+- The uncommitted dev-only files, by the user's decision.

docs/plans/2026-09-11-1526-provider-health-checks.md

@@ -0,0 +1,354 @@
+# Plan: Provider health checks — see whether the cluster can reach each provider
+**Created:** 2026-09-11 15:26 UTC
+
+## Context
+
+Today the operator can't say whether a provider's backend is reachable. The `Provider`
+interface ([provider.go:89-109](internal/provider/provider.go#L89-L109)) only has
+Create/Get/Delete/ListByTag, and `/healthz` and `/readyz` are bare `healthz.Ping`. A
+broken KVM path only shows up later: as a Proxy stuck in Provisioning, or as an orphan-GC
+sweep logging an error. Several failures stay invisible until a Proxy needs that host:
+
+- a missing `kvm-ssh` Secret;
+- a rotated host key;
+- a Tailscale route that is down;
+- a storage pool that is inactive or full.
+
+The goal is a periodic sanity check of every configured provider from inside the cluster,
+with results you can see through `kubectl` and Prometheus.
+
+Decisions made by the user during planning:
+
+1. **Surface:** a new operator-owned, read-only, cluster-scoped **`ProviderHealth`** CRD,
+   with one object per configured provider, a `Ready` condition and per-target detail.
+   Also Prometheus gauges, and a log line whenever a target goes up or down. No
+   EventRecorder, matching the earlier decision.
+2. **Scope:** real checks for all three providers.
+   - **kvm:** checked per host.
+   - **kubernetes:** confirms the API server and RBAC work.
+   - **gcp:** confirms credentials and the API work.
+3. **KVM depth:** reachability plus the storage pool.
+   - Over the existing pooled SSH+libvirt connection, call the libvirt version and
+     hostname RPCs.
+   - Confirm the configured pool is running and report its free space.
+   - No bridge check.
+
+Not affected: the Proxy reconcile behaviour, and the manager's `/healthz` and `/readyz`.
+A dead hypervisor must never make the manager unready.
+
+## Design
+
+### Contract: `internal/provider/check.go`
+
+```go
+type TargetStatus struct {
+    Name, Address   string
+    Ready           bool
+    Reason, Message string
+    Latency         time.Duration
+    Details         map[string]string
+}
+// Check never errors; each target carries its own verdict. Must return promptly once ctx ends.
+type Checker interface{ Check(ctx context.Context) []TargetStatus }
+```
+
+- **Shared reasons:** `Reachable`, `Timeout`, `Unreachable`, `Unauthorized`, `Forbidden`,
+  `APIError`.
+- **Compile-time checks:** each provider's `check.go` has `var _ provider.Checker = (*Provider)(nil)`.
+- **Bypass the wrappers.** `cmd/main.go:197-201` wraps every provider in
+  `WithTracing(WithMetrics(...))`, so the Checkers are collected from the **raw** providers
+  map, before that loop. As a result, checks don't inflate `provider_requests_total`.
+
+### The checks
+
+**kvm** (`internal/provider/kvm/check.go`)
+- **Concurrency:** one goroutine per `cfg.Hosts` entry (`wg.Go`), writing to `out[i]` so
+  results stay in config order.
+- **Inside `p.hosts.call`:** `ConnectGetLibVersion`, `ConnectGetHostname`,
+  `StoragePoolLookupByName(host.Pool)`, then `StoragePoolGetInfo`.
+- **Rules for `fn`:**
+  - It returns only raw RPC errors. `call` drops the shared connection on any other error
+    ([conn.go:160](internal/provider/kvm/conn.go#L160)), so "pool not running" is decided
+    after `call` returns, never returned from `fn`.
+  - It fills a captured result struct only after all four calls succeed. That struct is
+    read only when `call` returned nil, which avoids a race with an abandoned goroutine.
+- **Details:** `hostname`, `libvirtVersion` (decoded as major.minor.patch), `pool`,
+  `poolState`, `poolCapacity` and `poolAvailable` (in GiB).
+- **New RPCs:** add the three to `libvirtAPI` ([conn.go:29-47](internal/provider/kvm/conn.go#L29-L47)),
+  `tracingAPI` (`trace.go`) and `fakeAPI` (`fake_test.go`). `fakeAPI` gets a hostname, a
+  fixed version, and a per-pool state map that defaults to Running.
+- **Telling failure stages apart without matching strings:** add
+  `type dialError struct{ stage dialStage; err error }` in `conn.go`, with `Error()`
+  returning the inner message unchanged and `Unwrap()`. It wraps at four places:
+  - `lazySSH.dial`, when loading the key material fails (`ssh-material`)
+  - `dialSSH`, at the dial step (`tcp`)
+  - `dialSSH`, at the handshake (`ssh-handshake`)
+  - `connectLibvirt` (`libvirt-connect`)
+
+  Because `Unwrap` is kept, `classify` and `isClassified` still see `ErrPermanent` and
+  `fs.ErrNotExist`, so reconcile behaviour doesn't change.
+- **`checkReason(err)`, tested in order:**
+
+  | Condition | Reason |
+  |---|---|
+  | stage ssh-material | `SSHMaterialUnavailable` |
+  | stage tcp | `Unreachable` |
+  | stage ssh-handshake | `SSHHandshakeFailed` |
+  | stage libvirt-connect | `LibvirtConnectFailed` |
+  | ctx deadline or cancel | `Timeout` |
+  | `ErrNoStoragePool` | `PoolNotFound` |
+  | any other libvirt error | `LibvirtError` |
+  | anything else | `Unreachable` |
+  | success, pool not running | `PoolNotActive` |
+  | success | `Reachable` |
+- **Accepted behaviour:**
+  - A check that hits its timeout drops the shared connection. That already happens for
+    any cancelled reconcile context, and it only fires when the host isn't answering.
+  - The leader now keeps one SSH connection per host open from startup, with keepalives
+    every 15s. Before, the connection only opened with the first KVM reconcile.
+  - No forced periodic redial. `lazySSH` caches the key material forever, so a redial
+    would not pick up a rotated Secret anyway; that gap goes into the docs under
+    "Known gaps".
+
+**kubernetes** (`internal/provider/kubernetes/check.go`)
+- **Call:** `List` pods with the managed-label selector and `client.Limit(1)`,
+  cluster-wide. It uses the same uncached client, verb and RBAC as `ListByTag` and GC.
+  There is no namespace setting in the kubernetes config.
+- **Target:** one target named `apiserver`, with the address from a new `apiHost` field
+  set in `NewWithTransportWrapper`.
+- **Reasons:**
+  - `Timeout`: the context ended, or the error is `IsTimeout`/`IsServerTimeout`.
+  - `Unauthorized`, `Forbidden`.
+  - `APIError`: any other API status.
+  - `Unreachable`: anything else.
+
+**gcp** (`internal/provider/gcp/check.go`, plus a seam change in `gcp.go`)
+- **Why a new seam method:** the existing `AggregatedList` seam reads every page
+  ([gcp.go:63-78](internal/provider/gcp/gcp.go#L63-L78)), and MaxResults is only the page
+  size. So add `Probe(ctx, *computepb.AggregatedListInstancesRequest) error`, which makes
+  one `it.Next()` call and treats `iterator.Done` as nil.
+- **Request:** `Project`, `MaxResults: 1`, `ReturnPartialSuccess: true`, and the same
+  managed-label filter `ListByTag` uses.
+- **Target:** one per project, `Name: cfg.Project`, `Address: compute.googleapis.com`.
+  GCP config has no zones, and reading them from Proxy placements would tie the check to
+  the Proxy cache.
+- **Reasons:**
+  - `Timeout`: the context ended.
+  - For `*googleapi.Error`: 401 is `Unauthorized`, 403 is `Forbidden` (this also covers a
+    disabled API), anything else is `APIError`.
+  - `Unreachable`: anything else, including a failure to get a token.
+
+### The `ProviderHealth` CRD: `api/v1alpha1/providerhealth_types.go`
+
+- **Written by hand**, since kubebuilder isn't installed; `PROJECT` is edited by hand too.
+- **Scope and markers:** cluster-scoped, `shortName=ph`, status subresource, no spec.
+- **Scheme registration:** add the type to the existing `AddKnownTypes` call in
+  [proxy_types.go:362](api/v1alpha1/proxy_types.go#L362), not a second `init()`.
+- **Status:**
+  - `type`
+  - `conditions` (`Ready`; `ConditionReady` constant)
+  - `targets[]`, as a list-map keyed by `name`, each with: `name`, `address`, `ready`,
+    `reason`, `message` (at most 1024 characters), `latencyMillis`,
+    `lastTransitionTime`, `details`
+  - `targetsReady` (for example `"1/2"`, used by a printer column)
+  - `lastCheckTime`
+- **Printer columns:** Type, Ready, Targets, Last Check, Age.
+- **Viewer role:** `config/rbac/providerhealth_viewer_role.yaml`. No editor role, since the
+  operator owns the object.
+
+### Monitor: `internal/providerhealth`
+
+- **Fields:** `Monitor{Client, Reader, Targets []Target{Name, Type, Checker}, Interval, Timeout}`.
+- **Scheduling:**
+  - `NeedLeaderElection() true`.
+  - Checks once immediately, then on a ticker. Copy the ticker loop from
+    [gc.go:54-80](internal/gc/gc.go#L54-L80).
+- **Each cycle:**
+  - Runs every provider concurrently under `context.WithTimeout(Timeout)`, inside one
+    `tracing.Start` span.
+  - Stores the snapshot under a mutex; `Snapshot()` returns a deep copy.
+  - Logs changes. Going down logs `WARNING: provider target not ready` with provider,
+    target, address, reason and message. Coming back logs `Provider target ready`. A
+    target that is ready on first sight logs at V(1) only.
+- **Writes:**
+  - Get through `mgr.GetAPIReader()`, not the cache. That avoids a cluster-wide informer
+    and the watch permission it would need, whatever `--proxy-namespace` is set to.
+  - Create the object if it is missing, labelled
+    `app.kubernetes.io/managed-by=egress-proxies-operator`.
+  - Then `Status().Patch(MergeFrom(base))` when the status changed, following
+    [status.go:114-121](internal/controller/status.go#L114-L121).
+  - This means one write per provider per interval, because `lastCheckTime` moves each
+    time. That's acceptable given the 10s minimum interval.
+- **`buildStatus(prev, result, now)` is a pure function:**
+  - A target keeps its `lastTransitionTime` unless `ready` flipped.
+  - The `Ready` condition is set through `apimeta.SetStatusCondition`:
+    - `True/AllTargetsReady` when every target is ready
+    - `False/TargetsNotReady` otherwise, with the failing targets in the message
+    - `Unknown/ChecksNotImplemented` when the provider has no Checker
+    - `Unknown/NoTargets` when there are zero targets
+- **Cleanup, once at `Start`:** delete labelled objects whose name isn't in the config.
+  The config is loaded once per process, and unlabelled objects are never touched.
+- **Interval 0 (disabled):** run the cleanup with nothing kept, then return nil, so no
+  object is left showing a stale `Ready=True`.
+- **Invalid provider names:** provider names aren't validated as DNS-1123 names. For an
+  invalid name, warn once and skip the object writes; metrics are still emitted.
+- **`ValidateSchedule(interval, timeout)`** is called in `main.go` before the manager
+  starts. Interval must be 0 or at least 10s; timeout must be in (0, interval].
+- **RBAC markers on `Monitor`:**
+  - `providerhealths`: get, list, create, delete
+  - `providerhealths/status`: get, patch
+  - `make manifests` regenerates `config/rbac/role.yaml`.
+
+### Metrics and wiring
+
+- **New function:** `metrics.RegisterProviderTargets(reg, read func() []ProviderTarget) error`.
+  `Register`'s signature is left alone.
+- **Gauges:** a scrape-time collector, in the style of `constCollector`, reads the Monitor
+  snapshot and emits three gauges labelled `{provider,target}`:
+  - `proxy_operator_provider_target_up`
+  - `proxy_operator_provider_check_duration_seconds`
+  - `proxy_operator_provider_check_timestamp_seconds`
+- **Conversion closure:** the one that turns the Monitor snapshot into gauges lives in
+  `main.go`, next to `proxyPhaseCounts`. So `providerhealth` doesn't import `metrics`.
+- **No pool-bytes gauge:** that value is in status, and node-exporter is the right source
+  for disk alerts.
+- **`cmd/main.go`:**
+  - New flags `--provider-check-interval` (1m) and `--provider-check-timeout` (15s).
+  - Build the `Targets` from `cfg.Providers` in config order, with
+    `c, _ := raw[pc.Name].(provider.Checker)`.
+  - `mgr.Add(&providerhealth.Monitor{Client: mgr.GetClient(), Reader: mgr.GetAPIReader(), ...})`,
+    then register the collector.
+
+## Steps (one commit each, execution-log entry per step)
+
+**Step 0 — Save the plan.**
+- Create branch `feat/provider-health` off `main`.
+- Copy this plan to `docs/plans/2026-09-11-1526-provider-health-checks.md` and commit it
+  alone.
+- The uncommitted `.claude/settings.json` change stays out of every commit.
+
+**Step 1 — ProviderHealth API and CRD.**
+- `api/v1alpha1/providerhealth_types.go`
+- `zz_generated.deepcopy.go`, from `make generate`
+- `config/crd/bases/crawl.example.com_providerhealths.yaml`, from `make manifests`
+- the CRD and RBAC `kustomization.yaml` files, plus the viewer role
+- `PROJECT`
+
+**Step 2 — Contract plus the kubernetes and GCP checks.**
+- `internal/provider/check.go`
+- `internal/provider/kubernetes/{kubernetes.go,check.go,check_test.go}`
+- `internal/provider/gcp/{gcp.go,check.go,check_test.go,gcp_test.go}`
+
+**Step 3 — KVM check.**
+- `internal/provider/kvm/{conn.go,trace.go,check.go,check_test.go,fake_test.go,conn_test.go,trace_test.go,kvm_integration_test.go}`
+
+**Step 4 — Monitor.**
+- `internal/providerhealth/{monitor.go,status.go,monitor_test.go,status_test.go,envtest_test.go}`
+- `config/rbac/role.yaml`, from `make manifests`
+
+**Step 5 — Metrics and wiring.**
+- `internal/metrics/{metrics.go,metrics_test.go}`
+- `cmd/main.go`
+- `config/manager/manager.yaml`: a comment for the new flags
+- `test/e2e/e2e_test.go`
+
+**Step 6 — Docs.**
+- `docs/architecture.md`: a component row plus a provider-health section
+- `docs/kvm-provider.md`: the reason table, the shared-connection behaviour, and the
+  Secret-rotation gap under "Known gaps"
+- `docs/testing.md`: the new tests and the commands
+- `CLAUDE.md`: its Architecture and Gotchas sections still say "empty scaffold"; point
+  them at `docs/architecture.md`
+
+**Step 7 — Validate, record, push, open the PR.**
+- Run `go vet ./...`, `go test -short ./...`, `make test` and `make test-e2e`.
+- Hand the user the KVM commands below, and record their results in the execution log.
+- Add a CHANGELOG entry once the user confirms.
+- Push, using the token-file credential helper from memory.
+- Open the PR against `main` through the Gitea API (`tea` isn't installed).
+- Don't merge.
+
+## Tests
+
+**Provider checks** (table-driven, `Test<Func>_<scenario>`):
+- kubernetes:
+  - reachable
+  - list options: `Limit==1`, the label selector, no namespace
+  - failure reasons: forbidden, unauthorized, timeout, cancelled, internal, and a
+    network `*url.Error`
+- gcp:
+  - reachable, asserting the request fields
+  - failure reasons: 401, 403, 500, deadline, network
+  - `TestRealInstances_Probe_fetchesOnePage` against `httptest`: exactly one request with
+    `maxResults=1`, even though the page has a `nextPageToken`
+- kvm (using `hostFakes`/`newTestProvider` from `kvm_test.go`):
+  - reachable, with details filled in
+  - per-host verdicts, order preserved
+  - pool inactive, pool missing
+  - hosts checked concurrently (a sequential check would deadlock)
+  - a pool verdict keeps the connection (the dial count stays 1)
+  - a `checkReason` table
+  - `TestSSHAuth_dial_stages`, using the in-process SSH server in `conn_test.go`
+  - three new rows in the tracing table
+
+**Monitor:**
+- Setup: the fake client is built with a status subresource and a RESTMapper that
+  registers ProviderHealth as cluster-scoped.
+- Tests:
+  - `buildStatus` table
+  - objects are created with the label, and existing ones are patched
+  - the timeout bounds a checker that blocks
+  - a nil checker gives Unknown
+  - only transitions are logged (`funcr` logger)
+  - cleanup deletes only labelled objects that aren't configured
+  - the disabled path cleans up and returns
+  - an invalid name skips the writes
+  - `ValidateSchedule` table
+- An envtest round-trip, skipped under `-short`, which proves the generated schema is
+  accepted by a real apiserver.
+
+**Metrics:** `TestRegisterProviderTargets_scrape`, using a fresh registry and
+`GatherAndCompare`; an empty snapshot produces no series.
+
+**e2e (kind):** `providerhealth kubernetes` reaches `Ready=True`, which proves the
+generated RBAC works end to end.
+
+## Verification
+
+**Automated (Claude runs):** `go vet ./...`, `go test -short ./...`, `make test` (envtest
+with `-race`) and `make test-e2e`.
+
+**Against the real hypervisor (handed to the user, not run by Claude)**, with beelink as
+the host:
+
+```bash
+KVM_TEST_HOST=beelink.tail8c69c.ts.net:22 KVM_TEST_USER=novakj KVM_TEST_SSH_KEY=~/.ssh/id_ed25519 \
+KVM_TEST_KNOWN_HOSTS=~/.ssh/known_hosts KVM_TEST_BRIDGE=<bridge> KVM_TEST_IMAGE_URL=<image-url> \
+  go test -tags=integration -count=1 -v -run TestKVMCheck ./internal/provider/kvm/
+
+make install && go run ./cmd/main.go --providers-config <scratch copy of kvm-home config, local key paths> \
+  --metrics-bind-address :8080 --metrics-secure=false --provider-check-interval=15s
+kubectl get ph -w
+kubectl get ph kvm-home -o yaml
+curl -s localhost:8080/metrics | grep -E 'proxy_operator_provider_(target_up|check_)'
+```
+
+Each negative case is a one-line edit to the scratch config; nothing changes on the host.
+
+| Edit | Expected reason |
+|---|---|
+| `pool: nope` | `PoolNotFound` |
+| `address: …:2` | `Unreachable` |
+| `knownHostsFile: /dev/null` or `user: nobody` | `SSHHandshakeFailed` |
+| `sshKeyFile: /nonexistent` | `SSHMaterialUnavailable` |
+| `socket: /nonexistent` | `LibvirtConnectFailed` |
+
+## Out of scope
+
+- Gating Proxy creates on a host that is known to be down.
+- Tying `/readyz` to provider health.
+- Bridge checks.
+- SelfSubjectAccessReview checks of create/delete rights.
+- Hot-reloading the providers config.
+- Picking up a rotated `kvm-ssh` Secret without a restart. This is an existing gap; it
+  gets documented.

docs/plans/2026-09-16-0841-provider-blueprint.md

@@ -0,0 +1,119 @@
+# Plan: Provider blueprint (docs/providers.md)
+**Created:** 2026-09-16 08:41
+
+## Context
+
+The provider contract is spread across `docs/prompts/__initial-prompt.md` §5,
+the comments in `internal/provider/{provider,errors,check}.go`, `docs/architecture.md`
+§5 and §11, `docs/kvm-provider.md`, and plan Step 2/3/8. Many rules exist only in
+code: providerID must be self-contained, UID/CreatedAt must be read back, Create
+must survive a crash partway through, calls must stop when ctx is cancelled,
+decorators are wired in `cmd/main.go`, and so on. With "~5 providers soon", the
+next provider author needs one blueprint. It should say what a provider MUST,
+SHOULD or MAY do, show which features each existing provider (kubernetes, gcp,
+kvm) actually implements (with file:line evidence), give a checklist for adding
+one, and list the gaps.
+
+Decisions (user): store in `docs/providers.md`; scope = current state plus a
+separate "Gaps and proposed features" section (nothing implemented); wait for
+`feat/provider-health` to merge, then branch off `main`.
+
+## Step 0 — Save the plan; wait for the merge
+
+- Right after ExitPlanMode, copy this plan to `docs/plans/2026-09-16-0841-provider-blueprint.md`
+  and commit it on its own. The current branch is `feat/provider-health`, so the
+  plan reaches `main` with that merge. CLAUDE.md requires the plan commit to
+  come before any branching.
+- Stop and wait for the user to confirm that `feat/provider-health` is merged.
+  Then `git switch main && git pull`, and `git switch -c feat/provider-blueprint`.
+
+## Step 1 — Collect evidence (read-only, in parallel)
+
+Run 3 Explore agents at the same time, one each for `internal/provider/kubernetes`,
+`internal/provider/gcp` and `internal/provider/kvm`, including their tests. Each
+answers the **same questionnaire**, citing `file:line` for every answer and writing
+"not implemented" explicitly when that's the case:
+
+1. **Identity**: providerID format and parsing; what it carries (zone, host, namespace).
+2. **Create**: how it stays idempotent (name lookup, 409, DomainLookupByName);
+   whether it returns before the VM is up or waits; what happens when an earlier
+   attempt crashed partway; which Placement fields are used or ignored; how
+   cloud-init is delivered; how the port is used; how GC labels/UID are attached.
+3. **Get**: native state → InstanceState mapping; IP discovery rules; UID and
+   CreatedAt readback; how NotFound is detected.
+4. **Delete**: idempotency; what gets removed (disks, volumes, seeds); returns
+   before or after the resource is gone; unknown/malformed providerID behaviour.
+5. **ListByTag**: scope (namespaces, zones, hosts); partial-failure behaviour; how
+   managed resources are recognised.
+6. **Error classification**: table of backend error → NotFound/Quota/Transient/
+   Permanent; whether quota is detected at all.
+7. **Health check** (`Checker`): targets, provider-specific Reasons, Details keys.
+8. **Config**: YAML section, defaults, validation (`internal/provider/config*.go`),
+   credentials/Secrets, which settings are read once per process.
+9. **Connections and cancellation**: pooling, reconnects, timeouts, ctx honouring.
+10. **Observability**: log keys and verbosity levels, spans (`kvm/trace.go`, gcp
+    `wirelog.go`), anything beyond the shared metrics/tracing decorators.
+11. **Testing**: fakes, envtest/kind, `//go:build integration` tests, what is
+    not covered.
+12. **Known gaps**: TODOs, stale comments, documented limitations.
+
+Meanwhile, read the shared layer myself: `internal/provider/{provider,errors,check,
+name,config,metrics,tracing}.go`, `internal/provider/registry/registry.go`,
+`cmd/main.go` (registry wiring, decorator order, Checker collection), and the
+places that consume the contract, so every MUST rule traces to a caller that depends on it:
+`internal/controller` (error-class branching, ErrNotFound → replace/adopt),
+`internal/gc` (UID, CreatedAt age skip, ListByTag errors),
+`internal/providerhealth`. Then cross-check against the existing docs listed in Context.
+
+## Step 2 — Write `docs/providers.md`
+
+Sections:
+
+1. **What a provider is**: a few sentences plus links to architecture.md §5/§11.
+2. **The contract**: per method (Create/Get/Delete/ListByTag), MUST/SHOULD/MAY
+   rules, each followed by *why*: the reconciler or GC behaviour it protects,
+   linked to code.
+3. **Error taxonomy**: what each class makes the reconciler do; how to map backend
+   errors; "unclassified = transient" pitfall.
+4. **Identity and GC labels**: providerID guidance, `LabelManaged`/`LabelUID`,
+   CreatedAt, deterministic `NameFromUID`.
+5. **Optional interfaces**: `Checker` (shared reasons, never fails as a whole,
+   collected before decorators).
+6. **Feature matrix**: one row per questionnaire item, columns kubernetes/gcp/kvm,
+   cells `yes`/`partial`/`no` plus a short note. Put the file:line evidence in a
+   compact per-provider notes subsection under the table, not in the cells.
+7. **Adding a provider: checklist**: package layout, config section and validation,
+   registry constructor, `cmd/main.go` wiring, Checker, error mapping with tests,
+   `name`/label tests, integration test behind a build tag, provider doc page,
+   architecture.md rows, demo script (optional).
+8. **Gaps and proposed features**: current-state gaps per provider (e.g. no kvm
+   quota detection, orphan volumes invisible to GC, Secret rotation needs a
+   restart, ListByTag partial-failure inconsistency, stale `mock` references in
+   `provider.go` comments), plus proposals that are clearly marked as not
+   implemented. Only list them; don't design them.
+
+Also: link it from `docs/architecture.md` (Components table → provider contract row)
+and from the CLAUDE.md "Architecture" section. Leave per-provider docs where they are.
+
+Follow the plan-execution convention: create `docs/plans-executions/2026-09-16-0841-provider-blueprint.md`
+with the status checklist and step entries, and commit it with each step.
+
+## Step 3 — Verify, then open the MR
+
+- A verification agent (Explore) re-checks each matrix cell and each MUST rule
+  against the cited file:line and reports any mismatch. Fix what it finds.
+- Check that every relative link in the new and edited docs resolves (grep the
+  `](` targets and `ls` each one).
+- `go build ./... && go test -short ./...` as a sanity check (docs-only change,
+  expected no-op).
+- Push `feat/provider-blueprint` and open the MR against `main` via the Gitea API,
+  following the project's push/MR memory note. After the user confirms, add a CHANGELOG.md entry.
+
+## Critical files
+
+- New: `docs/providers.md`, `docs/plans/2026-09-16-0841-provider-blueprint.md`,
+  `docs/plans-executions/2026-09-16-0841-provider-blueprint.md`
+- Edited: `docs/architecture.md` (link), `CLAUDE.md` (link)
+- Read-only sources: `internal/provider/**`, `cmd/main.go`, `internal/controller`,
+  `internal/gc`, `internal/providerhealth`, `docs/prompts/__initial-prompt.md`,
+  `docs/kvm-provider.md`, `docs/gcp-vm-validation.md`

docs/testing.md

@@ -53,18 +53,23 @@ manual runs too.
 | | `config_test.go` | providers-config parsing + fail-fast validation | pure |
 | | `metrics_test.go` | `WithMetrics` decorator: classified result labels, error passthrough | fake recorder + static provider |
 | `internal/provider/kubernetes` | `kubernetes_test.go`, `pod_test.go` | Create/Get/Delete/ListByTag over real Pod objects; pure `buildPod`/`squidConf` incl. the `max_filedescriptors` OOM-regression assertion | `client/fake` (real `corev1.Pod`s, no kubelet) |
+| | `check_test.go` | health check: a single cluster-wide list with `Limit=1` and the managed-label selector; the reason for every API error class | `client/fake` + interceptor |
 | `internal/provider/gcp` | `insert_test.go`, `errors_test.go`, `gcp_test.go` | field-by-field `buildInsertRequest`; HTTP-code classification table; zone-qualified IDs, 409-is-success, 9-row state mapping, ListByTag filter + `ReturnPartialSuccess` | pure builder needs **no fake**; the rest uses the flattened `instancesAPI` seam |
+| | `check_test.go` | health check: request shape and the reason per HTTP code or network error; the **real SDK client** makes exactly one request with `maxResults=1`, even when a next page exists | fake seam + `httptest` |
 | `internal/provider/kvm` | `domainxml_test.go`, `volume_test.go`, `seed_test.go`, `id_test.go`, `metadata_test.go` | pure builders: domain/volume XML unmarshalled back and checked field by field, seed ISO reopened with the same library, providerID round-trip, metadata in both namespace spellings | pure |
 | | `errors_test.go`, `conn_test.go` | libvirt-code classification table; connection pool (reuse, redial, drop on transport error, cancel releases a blocked RPC, one dial per host under concurrency); SSH dial, host-key policy, keepalive | fake dialer; a **real in-process `x/crypto/ssh` server** forwarding a unix socket |
 | | `image_test.go`, `kvm_test.go` | image cache (fresh fetch, ready, marker missing, checksum change/mismatch, 4xx/5xx, concurrent callers); Create resume after every partial step, leftover truncated seed, state/IP mapping, every Delete already-gone case, ListByTag across hosts / one down / unreadable domain; logging tiers (V0 silent, V1 one line per RPC, V2 XML) with cloud-init never logged | `httptest` image server + in-memory `fakeAPI` behind the `libvirtAPI` seam |
 | | `trace_test.go`, `conn_test.go`, `errors_test.go` | every `libvirtAPI` method traced exactly once (guarded by reflection on the interface), results and errors pass through, libvirt error fields, detail keys only at V2, upload content never logged; pool connect/drop lines, no wrapping below V1 | `fakeAPI` + `funcr` capture |
-| | `kvm_integration_test.go` (`-tags=integration`) | one real VM: image fetch, Create, idempotent re-Create, Running with IP, ListByTag, Delete, pool state afterwards | a **real hypervisor** — see below |
+| | `check_test.go`, `conn_test.go` | health check: details, per-host verdicts in config order, hosts checked concurrently, a timeout that keeps the connection and refuses a second check while the first still waits, no waiting behind a reconcile's dial, the reason table; `probe` abandons without closing, but still drops on a transport error; waiting for a host's slot ends with the caller's deadline; an inactive or missing pool and daemon errors keep the connection; every dial stage tagged through the real SSH server, with classification unchanged | `fakeAPI`, fake dialers, in-process SSH server |
+| | `kvm_integration_test.go` (`-tags=integration`) | one real VM: image fetch, Create, idempotent re-Create, Running with IP, ListByTag, Delete, pool state afterwards; `TestKVMCheck_realHost` runs the health check against the host and creates nothing | a **real hypervisor** — see below |
 | `internal/health` | `probe_test.go` | probes through a **real CONNECT-capable proxy stub** to a real TLS server: tunnel success, refused CONNECT, wrong status, dead proxy, plain-http forward | `httptest` + hijacked bidirectional tunnel |
 | | `engine_test.go` | thresholds, first-verdict, latency suppression matrix, dropped-event retry, stale-probe UID guard, tick scheduling/seeding/pruning, `Start` end-to-end | fake `client.Reader` + injected `probeFn` |
 | `internal/lease` | `store_test.go` | capacity, `MaxLeases=0`, all three selection tie-breaks, target-scoped vs global cooldowns, expiry via fake clock, report-on-retained-lease, idempotent release, **40 concurrent acquires never exceeding capacity** | injectable clock; the concurrency case is why `-race` matters |
 | `internal/discovery` | `server_test.go` | auth matrix (incl. `/healthz` exemption), list filtering, grant shape, full 409 arithmetic, invalid TTL/body/result, idempotent release, report→cooldown→409 round trip, `Start`/shutdown | `httptest` over the real handler chain, fake cache reader, **real** `lease.Store` |
 | `internal/gc` | `gc_test.go` | the true-orphan matrix (live/deleting/young/unlabelled all kept), broken-provider isolation, list-failure skips sweep, namespace guard, sweep loop | fake reader + canned-list provider |
-| `internal/metrics` | `metrics_test.go` | scrape-time collectors (`GatherAndCompare`), per-proxy series cleanup via `ForgetProxy`, label counts, fresh-registry-per-test property | `prometheus/client_golang/testutil` |
+| `internal/providerhealth` | `status_test.go`, `monitor_test.go` | `buildStatus` (the Ready condition, per-target transition times, truncation); objects created with the label and patched, the round timeout, no-checker Unknown, invalid names, transition-only logging, cleanup of labelled unconfigured objects, the checks-off path, `ValidateSchedule` | fake client with a cluster-scoped RESTMapper + `funcr` capture |
+| | `envtest_test.go` | the generated `ProviderHealth` schema against a real API server: cluster scope, status subresource, required fields, list-map keys, the 1024-character message limit | envtest (skipped under `-short`) |
+| `internal/metrics` | `metrics_test.go` | scrape-time collectors (`GatherAndCompare`) incl. the provider-target gauges, per-proxy series cleanup via `ForgetProxy`, label counts, fresh-registry-per-test property | `prometheus/client_golang/testutil` |
 
 Conventions (from `CLAUDE.md`): stdlib `testing`, table-driven by default,
 `t.Parallel()` where safe, tests next to source, ginkgo only in the
@@ -89,7 +94,7 @@ envtest suite the scaffold generated.
 - **Live GCP**: no test talks to Google. The seam boundary
   (`buildInsertRequest` + classification) is tested exhaustively instead.
 - **Live libvirt, by default**: the KVM unit tests run over an in-memory
-  fake of the 17 libvirt calls the provider uses. Whether libvirt accepts
+  fake of the 20 libvirt calls the provider uses. Whether libvirt accepts
   the domain XML, whether a raw qcow2 upload plus pool refresh yields a
   usable backing image, and whether the guest actually boots and reports
   an address are provable only on a real host — that is the opt-in
@@ -109,13 +114,19 @@ had.
 export KVM_TEST_HOST=nas1.home.lan:22          # SSH endpoint
 export KVM_TEST_USER=virt                      # must be in the libvirt group
 export KVM_TEST_SSH_KEY=$HOME/.ssh/id_ed25519  # private key for that user
+export KVM_TEST_KNOWN_HOSTS=$HOME/.ssh/kvm_known_hosts  # the hypervisors' host keys
 export KVM_TEST_BRIDGE=br0                     # bridge to a network with DHCP
 export KVM_TEST_IMAGE_URL=https://cloud.debian.org/images/cloud/trixie/latest/debian-13-generic-amd64.qcow2
 make test-integration-kvm
 ```
 
-Optional: `KVM_TEST_KNOWN_HOSTS` (without it host keys are **not**
-verified — the log says so), `KVM_TEST_IMAGE_SHA256`, `KVM_TEST_POOL`
+`KVM_TEST_KNOWN_HOSTS` is technically optional, but leaving it out turns
+host-key verification **off** for the run (the log says so); keep a
+`kvm_known_hosts` holding just the hypervisors rather than pointing the
+test at your everyday `known_hosts`, so a re-keyed hypervisor is a
+one-line fix in a file nothing else uses.
+
+Also optional: `KVM_TEST_IMAGE_SHA256`, `KVM_TEST_POOL`
 (default `default`), `KVM_TEST_SOCKET`, `KVM_TEST_TIMEOUT` (default
 `20m`, must cover the first-time image download), `KVM_TEST_KEEP`
 to leave the VM behind for `virsh console` when something goes wrong,
@@ -139,13 +150,24 @@ but no per-VM overlay or seed. It also tries a TCP connect to port 22 on
 the guest, but only logs the outcome — the machine running the test may
 not route to the bridge network.
 
+`TestKVMCheck_realHost` runs the provider health check
+([kvm-provider.md](kvm-provider.md#health-check)) against the same host.
+It must find the host reachable and the pool running, and it creates
+nothing. It needs the same variables; to run only this test:
+
+```sh
+go test -tags=integration -count=1 -v -run TestKVMCheck ./internal/provider/kvm/
+```
+
 The test is not part of `make test` and not run in CI. Keep it compiling
 with `go vet -tags=integration ./...` when the provider's internals move.
 
 ## The scaffolded `make test-e2e` suite
 
 `test/e2e/` is the kubebuilder-generated smoke suite (build image → kind
-cluster → deploy → assert the manager pod runs and serves metrics). It
+cluster → deploy → assert the manager pod runs, publishes a Ready
+`ProviderHealth` for the `kubernetes` provider, and serves metrics that
+include that provider's `proxy_operator_provider_target_up` series). It
 compiles only under `-tags=e2e`, manages its own kind cluster
 (`make test-e2e` / `make cleanup-test-e2e`), and has been kept compiling
 (`go vet -tags=e2e ./...` is part of the routine) but is **not part of

internal/controller/proxy_controller.go

@@ -270,8 +270,14 @@ func (r *ProxyReconciler) reconcileDelete(ctx context.Context, p *crawlv1alpha1.
 	if p.Status.ProviderID == "" {
 		// Nothing was ever recorded as created; orphan GC reaps any stray
 		// instance a crashed create might have left behind.
-		controllerutil.RemoveFinalizer(p, crawlv1alpha1.FinalizerName)
-		return ctrl.Result{}, r.Update(ctx, p)
+		return r.releaseFinalizer(ctx, p)
+	}
+	if p.Annotations[crawlv1alpha1.AnnotationSkipCleanup] == "true" {
+		// Checked before the provider lookup so it also releases a Proxy whose
+		// provider is gone from the config. The instance is left running.
+		logf.FromContext(ctx).Info("skip-cleanup annotation set; abandoning the instance",
+			"providerID", p.Status.ProviderID, "provider", p.Spec.Provider)
+		return r.releaseFinalizer(ctx, p)
 	}
 	prov, ok := r.Providers[p.Spec.Provider]
 	if !ok {
@@ -279,11 +285,22 @@ func (r *ProxyReconciler) reconcileDelete(ctx context.Context, p *crawlv1alpha1.
 			"provider %q is not configured; cannot clean up instance %s", p.Spec.Provider, p.Status.ProviderID)
 	}
 	_, err := prov.Get(ctx, p.Status.ProviderID)
-	if provider.Class(err) == provider.ErrNotFound {
-		controllerutil.RemoveFinalizer(p, crawlv1alpha1.FinalizerName)
-		return ctrl.Result{}, r.Update(ctx, p)
+	switch {
+	case err == nil:
+	case provider.Class(err) == provider.ErrNotFound:
+		return r.releaseFinalizer(ctx, p)
+	case provider.Class(err) == provider.ErrPermanent:
+		// Get will never answer for this instance — a host retired from the
+		// config, a providerID the provider can no longer resolve. Delete is
+		// idempotent and may still finish, so try it instead of polling a Get
+		// that cannot converge, and release the object when it does.
+		if delErr := prov.Delete(ctx, p.Status.ProviderID); delErr != nil {
+			return r.deletionFailure(errors.Join(err, delErr))
 		}
-	if err != nil {
+		logf.FromContext(ctx).Info("instance unreachable but cleanup succeeded; removing the finalizer",
+			"providerID", p.Status.ProviderID, "error", err.Error())
+		return r.releaseFinalizer(ctx, p)
+	default:
 		return r.deletionFailure(err)
 	}
 	if err := prov.Delete(ctx, p.Status.ProviderID); err != nil {
@@ -294,6 +311,11 @@ func (r *ProxyReconciler) reconcileDelete(ctx context.Context, p *crawlv1alpha1.
 	return ctrl.Result{RequeueAfter: r.DeletionPoll}, nil
 }
 
+func (r *ProxyReconciler) releaseFinalizer(ctx context.Context, p *crawlv1alpha1.Proxy) (ctrl.Result, error) {
+	controllerutil.RemoveFinalizer(p, crawlv1alpha1.FinalizerName)
+	return ctrl.Result{}, r.Update(ctx, p)
+}
+
 func (r *ProxyReconciler) reconcileExternal(_ context.Context, p *crawlv1alpha1.Proxy) (ctrl.Result, error) {
 	if p.Spec.Endpoint == nil {
 		// CEL guarantees an endpoint on any object that went through the API

internal/controller/reconcile_test.go

@@ -2,6 +2,7 @@ package controller
 
 import (
 	"context"
+	"slices"
 	"testing"
 	"time"
 
@@ -78,6 +79,10 @@ func notFoundErr() error {
 	return provider.Wrap(provider.ErrNotFound, "get", "stub", "some-id", nil)
 }
 
+func permanentErr() error {
+	return provider.Wrap(provider.ErrPermanent, "get", "stub", "some-id", nil)
+}
+
 func testScheme(t *testing.T) *runtime.Scheme {
 	t.Helper()
 	s := runtime.NewScheme()
@@ -125,6 +130,15 @@ func withSpecHashAnnotation(hash string) func(*crawlv1alpha1.Proxy) {
 	}
 }
 
+func withSkipCleanup() func(*crawlv1alpha1.Proxy) {
+	return func(p *crawlv1alpha1.Proxy) {
+		if p.Annotations == nil {
+			p.Annotations = map[string]string{}
+		}
+		p.Annotations[crawlv1alpha1.AnnotationSkipCleanup] = "true"
+	}
+}
+
 func deleting() func(*crawlv1alpha1.Proxy) {
 	return func(p *crawlv1alpha1.Proxy) {
 		now := metav1.Now()
@@ -375,6 +389,59 @@ func TestReconcile_actionTable(t *testing.T) {
 				assertProxyGone(t, r)
 			},
 		},
+		{
+			// A permanent Get (a retired KVM host, a providerID the provider
+			// cannot resolve) used to retry forever with the finalizer on.
+			name:       "deletion with an unreachable instance still deletes and releases",
+			proxy:      managedProxy(deleting(), withProviderID("stub-id-1")),
+			stub:       &stubProvider{getErr: permanentErr()},
+			wantResult: ctrl.Result{},
+			verify: func(t *testing.T, r *ProxyReconciler, stub *stubProvider) {
+				assertProxyGone(t, r)
+				if stub.deleteCalls != 1 {
+					t.Errorf("deleteCalls = %d, want 1 — Delete must be tried when Get cannot answer", stub.deleteCalls)
+				}
+			},
+		},
+		{
+			name:  "deletion keeps the finalizer when cleanup itself fails",
+			proxy: managedProxy(deleting(), withProviderID("stub-id-1")),
+			stub: &stubProvider{
+				getErr:    permanentErr(),
+				deleteErr: provider.Wrap(provider.ErrPermanent, "delete", "stub", "stub-id-1", nil),
+			},
+			wantResult: ctrl.Result{},
+			wantErr:    true,
+			verify: func(t *testing.T, r *ProxyReconciler, _ *stubProvider) {
+				// getProxy fails the test if the object is gone, which is what
+				// finalizer removal would have caused.
+				if p := getProxy(t, r); !slices.Contains(p.Finalizers, crawlv1alpha1.FinalizerName) {
+					t.Error("finalizer removed although cleanup failed; the instance would leak silently")
+				}
+			},
+		},
+		{
+			name:       "skip-cleanup annotation abandons the instance",
+			proxy:      managedProxy(deleting(), withProviderID("stub-id-1"), withSkipCleanup()),
+			stub:       &stubProvider{getInst: &provider.Instance{ID: "stub-id-1", State: provider.StateRunning}},
+			wantResult: ctrl.Result{},
+			verify: func(t *testing.T, r *ProxyReconciler, stub *stubProvider) {
+				assertProxyGone(t, r)
+				if stub.getCalls != 0 || stub.deleteCalls != 0 {
+					t.Errorf("getCalls/deleteCalls = %d/%d, want 0/0 — the provider must not be asked", stub.getCalls, stub.deleteCalls)
+				}
+			},
+		},
+		{
+			name: "skip-cleanup annotation releases a proxy whose provider is gone",
+			proxy: managedProxy(deleting(), withProviderID("stub-id-1"), withSkipCleanup(),
+				func(p *crawlv1alpha1.Proxy) { p.Spec.Provider = "retired" }),
+			stub:       &stubProvider{},
+			wantResult: ctrl.Result{},
+			verify: func(t *testing.T, r *ProxyReconciler, _ *stubProvider) {
+				assertProxyGone(t, r)
+			},
+		},
 		{
 			name:  "deletion deletes the instance and polls",
 			proxy: managedProxy(deleting(), withProviderID("stub-id-1")),

internal/metrics/metrics.go

@@ -117,3 +117,51 @@ func (c *constCollector) Collect(ch chan<- prometheus.Metric) {
 		ch <- prometheus.MustNewConstMetric(c.desc, prometheus.GaugeValue, float64(value), label)
 	}
 }
+
+// ProviderTarget is one provider health-check target as the gauges see it.
+type ProviderTarget struct {
+	Provider, Target string
+	Up               bool
+	Duration         time.Duration
+	CheckedAt        time.Time
+}
+
+// RegisterProviderTargets registers the provider-check gauges. Like the
+// other scrape-time collectors they are read from the source at every
+// scrape, so a target that goes away stops being reported rather than
+// leaving a stale series behind.
+func RegisterProviderTargets(reg prometheus.Registerer, read func() []ProviderTarget) error {
+	labels := []string{"provider", "target"}
+	return reg.Register(&providerTargetCollector{
+		up: prometheus.NewDesc("proxy_operator_provider_target_up",
+			"Whether the target passed its latest provider health check (1) or not (0).", labels, nil),
+		duration: prometheus.NewDesc("proxy_operator_provider_check_duration_seconds",
+			"Duration of the target's latest provider health check.", labels, nil),
+		checked: prometheus.NewDesc("proxy_operator_provider_check_timestamp_seconds",
+			"Unix time of the target's latest provider health check.", labels, nil),
+		read: read,
+	})
+}
+
+type providerTargetCollector struct {
+	up, duration, checked *prometheus.Desc
+	read                  func() []ProviderTarget
+}
+
+func (c *providerTargetCollector) Describe(ch chan<- *prometheus.Desc) {
+	ch <- c.up
+	ch <- c.duration
+	ch <- c.checked
+}
+
+func (c *providerTargetCollector) Collect(ch chan<- prometheus.Metric) {
+	for _, t := range c.read() {
+		up := 0.0
+		if t.Up {
+			up = 1
+		}
+		ch <- prometheus.MustNewConstMetric(c.up, prometheus.GaugeValue, up, t.Provider, t.Target)
+		ch <- prometheus.MustNewConstMetric(c.duration, prometheus.GaugeValue, t.Duration.Seconds(), t.Provider, t.Target)
+		ch <- prometheus.MustNewConstMetric(c.checked, prometheus.GaugeValue, float64(t.CheckedAt.UnixNano())/1e9, t.Provider, t.Target)
+	}
+}

internal/metrics/metrics_test.go

@@ -111,3 +111,45 @@ func TestRegister_freshRegistryPerTest(t *testing.T) {
 	register(t, m1, nil, 0)
 	register(t, m2, nil, 0)
 }
+
+func TestRegisterProviderTargets_scrape(t *testing.T) {
+	t.Parallel()
+	checked := time.Unix(1757602800, 0)
+	targets := []ProviderTarget{
+		{Provider: "kvm-home", Target: "nas1", Up: true, Duration: 84 * time.Millisecond, CheckedAt: checked},
+		{Provider: "kvm-home", Target: "nas2", Up: false, Duration: 15 * time.Second, CheckedAt: checked},
+	}
+	reg := prometheus.NewRegistry()
+	if err := RegisterProviderTargets(reg, func() []ProviderTarget { return targets }); err != nil {
+		t.Fatalf("RegisterProviderTargets: %v", err)
+	}
+
+	expected := `
+# HELP proxy_operator_provider_check_duration_seconds Duration of the target's latest provider health check.
+# TYPE proxy_operator_provider_check_duration_seconds gauge
+proxy_operator_provider_check_duration_seconds{provider="kvm-home",target="nas1"} 0.084
+proxy_operator_provider_check_duration_seconds{provider="kvm-home",target="nas2"} 15
+# HELP proxy_operator_provider_check_timestamp_seconds Unix time of the target's latest provider health check.
+# TYPE proxy_operator_provider_check_timestamp_seconds gauge
+proxy_operator_provider_check_timestamp_seconds{provider="kvm-home",target="nas1"} 1.7576028e+09
+proxy_operator_provider_check_timestamp_seconds{provider="kvm-home",target="nas2"} 1.7576028e+09
+# HELP proxy_operator_provider_target_up Whether the target passed its latest provider health check (1) or not (0).
+# TYPE proxy_operator_provider_target_up gauge
+proxy_operator_provider_target_up{provider="kvm-home",target="nas1"} 1
+proxy_operator_provider_target_up{provider="kvm-home",target="nas2"} 0
+`
+	if err := testutil.GatherAndCompare(reg, strings.NewReader(expected)); err != nil {
+		t.Error(err)
+	}
+}
+
+func TestRegisterProviderTargets_emptySnapshotHasNoSeries(t *testing.T) {
+	t.Parallel()
+	reg := prometheus.NewRegistry()
+	if err := RegisterProviderTargets(reg, func() []ProviderTarget { return nil }); err != nil {
+		t.Fatalf("RegisterProviderTargets: %v", err)
+	}
+	if n, err := testutil.GatherAndCount(reg); err != nil || n != 0 {
+		t.Errorf("series before the first check = %d (%v), want 0", n, err)
+	}
+}

internal/provider/check.go

@@ -0,0 +1,46 @@
+package provider
+
+import (
+	"context"
+	"time"
+)
+
+// Reasons shared by every Checker, so a listing reads the same across
+// providers. Providers add their own for failures only they can tell
+// apart (KVM's SSHHandshakeFailed or PoolNotActive, say).
+const (
+	ReasonReachable    = "Reachable"
+	ReasonTimeout      = "Timeout"
+	ReasonUnreachable  = "Unreachable"
+	ReasonUnauthorized = "Unauthorized"
+	ReasonForbidden    = "Forbidden"
+	ReasonAPIError     = "APIError"
+)
+
+// TargetStatus is one target's verdict from a Checker.
+type TargetStatus struct {
+	// Name identifies the target within its provider: a KVM host name,
+	// "apiserver", a GCP project.
+	Name string
+	// Address is what the check connected to, for humans.
+	Address string
+	Ready   bool
+	// Reason is ReasonReachable when Ready, otherwise why not.
+	Reason string
+	// Message is human-readable detail, usually the error.
+	Message string
+	Latency time.Duration
+	// Details are provider-specific facts gathered by the check.
+	Details map[string]string
+}
+
+// Checker is the optional interface of a provider that can verify its
+// backend is reachable and usable without touching any instance. The
+// provider health monitor type-asserts for it on the unwrapped providers,
+// so it stays out of the Provider contract.
+type Checker interface {
+	// Check returns one verdict per target, in a stable order. It never
+	// fails as a whole — a failure is a verdict — and must return promptly
+	// once ctx ends.
+	Check(ctx context.Context) []TargetStatus
+}

internal/provider/config_kvm.go

@@ -22,10 +22,18 @@ type KVMConfig struct {
 	// DiskSizeGB is the capacity of each VM's overlay disk. Default 10. The
 	// guest's cloud-init growpart expands the root filesystem into it.
 	DiskSizeGB int64 `json:"diskSizeGb,omitempty"`
-	// VendorData is cloud-init vendor-data placed on every seed next to the
-	// Proxy's own user-data. Default: installs and starts qemu-guest-agent so
-	// the provider can read the VM's IP even before it shows up in ARP.
+	// VendorData is optional cloud-init vendor-data placed on every seed next
+	// to the Proxy's own user-data. Note that cloud-init drops any top-level
+	// key of a vendor-data cloud-config that the user-data also defines.
+	// When InstallGuestAgent is on, this document and the agent script are
+	// carried as separate parts of a MIME multipart vendor-data.
 	VendorData string `json:"vendorData,omitempty"`
+	// InstallGuestAgent adds a vendor-data script that installs and starts
+	// qemu-guest-agent, which is how the provider learns a VM's address
+	// before the host's ARP table has seen traffic from it. Default true.
+	// Turn it off for images that already ship the agent, or that cannot
+	// reach a package mirror; discovery then depends on ARP alone.
+	InstallGuestAgent *bool `json:"installGuestAgent,omitempty"`
 	// Images are named base images; placement.image names a key here (or is
 	// an http(s) URL directly). The provider downloads each image into a
 	// host's storage pool on first use and caches it there.
@@ -80,8 +88,31 @@ type KVMHost struct {
 	Bridge string `json:"bridge"`
 }
 
+// Accepted KVMImage.Format and KVMImage.Firmware values.
+const (
+	KVMFormatQcow2  = "qcow2"
+	KVMFormatRaw    = "raw"
+	KVMFirmwareUEFI = "uefi"
+	KVMFirmwareBIOS = "bios"
+)
+
+// IsHTTPURL reports whether s is an absolute http(s) URL with a host, the
+// only kind of image source the kvm provider can fetch.
+func IsHTTPURL(s string) bool {
+	u, err := url.Parse(s)
+	return err == nil && (u.Scheme == "http" || u.Scheme == "https") && u.Host != ""
+}
+
 var sha256Hex = regexp.MustCompile(`^[0-9a-fA-F]{64}$`)
 
+// kvmIdentifier constrains the config names that are pasted into
+// identifiers elsewhere: a host name goes into the providerID
+// hosts/<name>/domains/<vm>, which is split on "/" to parse, and an image
+// key becomes the file name of the cached base volume. A slash in either
+// produces an ID or a path that never resolves again, so the VM leaks with
+// nothing able to address it.
+var kvmIdentifier = regexp.MustCompile(`^[A-Za-z0-9][A-Za-z0-9_.-]*$`)
+
 func validateKVM(prefix string, c *KVMConfig) error {
 	if c == nil {
 		return fmt.Errorf("%s: kvm block is required", prefix)
@@ -97,18 +128,21 @@ func validateKVM(prefix string, c *KVMConfig) error {
 	}
 	for name, img := range c.Images {
 		p := fmt.Sprintf("%s: kvm.images[%q]", prefix, name)
-		u, err := url.Parse(img.URL)
-		if err != nil || (u.Scheme != "http" && u.Scheme != "https") || u.Host == "" {
+		if !kvmIdentifier.MatchString(name) {
+			return fmt.Errorf("%s: key must be letters, digits, dots, dashes or underscores"+
+				" — it becomes the cached volume name base-<key>.<format>", p)
+		}
+		if !IsHTTPURL(img.URL) {
 			return fmt.Errorf("%s.url must be an http(s) URL (got %q)", p, img.URL)
 		}
 		if img.SHA256 != "" && !sha256Hex.MatchString(img.SHA256) {
 			return fmt.Errorf("%s.sha256 must be 64 hex characters", p)
 		}
-		if img.Format != "" && img.Format != "qcow2" && img.Format != "raw" {
-			return fmt.Errorf("%s.format must be qcow2 or raw (got %q)", p, img.Format)
+		if img.Format != "" && img.Format != KVMFormatQcow2 && img.Format != KVMFormatRaw {
+			return fmt.Errorf("%s.format must be %s or %s (got %q)", p, KVMFormatQcow2, KVMFormatRaw, img.Format)
 		}
-		if img.Firmware != "" && img.Firmware != "uefi" && img.Firmware != "bios" {
-			return fmt.Errorf("%s.firmware must be uefi or bios (got %q)", p, img.Firmware)
+		if img.Firmware != "" && img.Firmware != KVMFirmwareUEFI && img.Firmware != KVMFirmwareBIOS {
+			return fmt.Errorf("%s.firmware must be %s or %s (got %q)", p, KVMFirmwareUEFI, KVMFirmwareBIOS, img.Firmware)
 		}
 	}
 	if len(c.MachineTypes) == 0 {
@@ -128,6 +162,10 @@ func validateKVM(prefix string, c *KVMConfig) error {
 		if h.Name == "" {
 			return fmt.Errorf("%s: name is required", p)
 		}
+		if !kvmIdentifier.MatchString(h.Name) {
+			return fmt.Errorf("%s: name %q must be letters, digits, dots, dashes or underscores"+
+				" — it becomes part of the providerID hosts/<name>/domains/<vm>", p, h.Name)
+		}
 		if seen[h.Name] {
 			return fmt.Errorf("%s: duplicate host name %q", p, h.Name)
 		}

internal/provider/config_kvm_test.go

@@ -129,6 +129,19 @@ hosts:
 			yaml:       minimal("sshKeyFile: /key\nknownHostsFile: /kh\nmachineTypes:\n  small: {vcpus: 1, memoryMiB: 1024}\nhosts:\n  - {name: nas1, address: a:22, user: virt, bridge: br0}\n  - {name: nas1, address: b:22, user: virt, bridge: br0}\n"),
 			wantErrSub: `duplicate host name "nas1"`,
 		},
+		{
+			// A slash would produce hosts/dc1/rack2/domains/<vm>, which
+			// parseProviderID can never split back, so the VM would be
+			// unreachable for Get, Delete and GC alike.
+			name:       "host name with a slash",
+			yaml:       minimal("sshKeyFile: /key\nknownHostsFile: /kh\nmachineTypes:\n  small: {vcpus: 1, memoryMiB: 1024}\nhosts:\n  - {name: dc1/rack2, address: a:22, user: virt, bridge: br0}\n"),
+			wantErrSub: `name "dc1/rack2" must be letters, digits, dots, dashes or underscores`,
+		},
+		{
+			name:       "image key with a slash",
+			yaml:       minimal("sshKeyFile: /key\nknownHostsFile: /kh\nimages:\n  ../etc/d: {url: https://x/y.qcow2}\n" + hostsAndTypes),
+			wantErrSub: `images["../etc/d"]: key must be letters, digits, dots, dashes or underscores`,
+		},
 		{
 			name:       "host missing address",
 			yaml:       minimal("sshKeyFile: /key\nknownHostsFile: /kh\nmachineTypes:\n  small: {vcpus: 1, memoryMiB: 1024}\nhosts:\n  - {name: nas1, user: virt, bridge: br0}\n"),

internal/provider/gcp/check.go

@@ -0,0 +1,78 @@
+package gcp
+
+import (
+	"context"
+	"errors"
+	"net/http"
+	"time"
+
+	"cloud.google.com/go/compute/apiv1/computepb"
+	"google.golang.org/api/googleapi"
+	"google.golang.org/api/iterator"
+	"google.golang.org/protobuf/proto"
+
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+var _ provider.Checker = (*Provider)(nil)
+
+// computeAddress is the endpoint the SDK talks to, shown only to humans.
+const computeAddress = "compute.googleapis.com"
+
+// managedFilter selects the instances this operator created.
+const managedFilter = "labels." + provider.LabelManaged + " = " + provider.LabelManagedYes
+
+// Check asks the project for one page holding at most one managed
+// instance. That exercises the credentials, the Compute API and the
+// compute.instances.list permission orphan GC depends on, in a single
+// request whatever the project's size. The target is the project: the
+// provider config names no zones.
+func (p *Provider) Check(ctx context.Context) []provider.TargetStatus {
+	start := time.Now()
+	err := p.api.Probe(ctx, &computepb.AggregatedListInstancesRequest{
+		Project:              p.cfg.Project,
+		Filter:               proto.String(managedFilter),
+		MaxResults:           proto.Uint32(1),
+		ReturnPartialSuccess: proto.Bool(true),
+	})
+	ts := provider.TargetStatus{Name: p.cfg.Project, Address: computeAddress, Latency: time.Since(start)}
+	if err != nil {
+		logAPIError(p.logger(ctx), "check", err)
+		ts.Reason = checkReason(ctx, err)
+		ts.Message = err.Error()
+		return []provider.TargetStatus{ts}
+	}
+	ts.Ready = true
+	ts.Reason = provider.ReasonReachable
+	return []provider.TargetStatus{ts}
+}
+
+// Probe fetches the first page of an aggregated list and discards it. The
+// SDK treats MaxResults as the page size, so AggregatedList, which drains
+// the iterator, would still walk every page of the project.
+func (r *realInstances) Probe(ctx context.Context, req *computepb.AggregatedListInstancesRequest) error {
+	_, err := r.client.AggregatedList(ctx, req).Next()
+	if err == iterator.Done {
+		return nil
+	}
+	return err
+}
+
+// checkReason maps a failed probe to a reason. A token that cannot be
+// obtained never reaches the API, so it has no status code and lands on
+// Unreachable, with the detail in the message.
+func checkReason(ctx context.Context, err error) string {
+	var gerr *googleapi.Error
+	switch {
+	case ctx.Err() != nil, errors.Is(err, context.DeadlineExceeded):
+		return provider.ReasonTimeout
+	case !errors.As(err, &gerr):
+		return provider.ReasonUnreachable
+	case gerr.Code == http.StatusUnauthorized:
+		return provider.ReasonUnauthorized
+	case gerr.Code == http.StatusForbidden:
+		return provider.ReasonForbidden
+	default:
+		return provider.ReasonAPIError
+	}
+}

internal/provider/gcp/check_test.go

@@ -0,0 +1,165 @@
+package gcp
+
+import (
+	"context"
+	"errors"
+	"net/http"
+	"net/http/httptest"
+	"net/url"
+	"sync"
+	"syscall"
+	"testing"
+
+	compute "cloud.google.com/go/compute/apiv1"
+	"cloud.google.com/go/compute/apiv1/computepb"
+	"google.golang.org/api/googleapi"
+	"google.golang.org/api/option"
+	"google.golang.org/protobuf/proto"
+
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+func TestCheck_reachable(t *testing.T) {
+	t.Parallel()
+	api := &fakeAPI{}
+	p := newTestProvider(api)
+
+	got := p.Check(context.Background())
+
+	if len(got) != 1 {
+		t.Fatalf("Check() returned %d targets, want 1", len(got))
+	}
+	ts := got[0]
+	if !ts.Ready || ts.Reason != provider.ReasonReachable || ts.Message != "" {
+		t.Errorf("Check() = %+v, want ready and Reachable", ts)
+	}
+	if ts.Name != "my-project" || ts.Address != "compute.googleapis.com" {
+		t.Errorf("target = %q at %q, want my-project at compute.googleapis.com", ts.Name, ts.Address)
+	}
+	want := &computepb.AggregatedListInstancesRequest{
+		Project:              "my-project",
+		Filter:               proto.String("labels.proxy-operator-managed = true"),
+		MaxResults:           proto.Uint32(1),
+		ReturnPartialSuccess: proto.Bool(true),
+	}
+	if !proto.Equal(api.probeReq, want) {
+		t.Errorf("probe request = %v, want %v", api.probeReq, want)
+	}
+}
+
+func TestCheck_failureReasons(t *testing.T) {
+	t.Parallel()
+	ended, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	tests := []struct {
+		name string
+		ctx  context.Context
+		err  error
+		want string
+	}{
+		{"unauthorized", context.Background(), gerr(401), provider.ReasonUnauthorized},
+		{"forbidden", context.Background(), gerr(403, "forbidden"), provider.ReasonForbidden},
+		{"api disabled", context.Background(), gerr(403, "accessNotConfigured"), provider.ReasonForbidden},
+		{"server error", context.Background(), gerr(500), provider.ReasonAPIError},
+		{"deadline", context.Background(), &url.Error{
+			Op: "Get", URL: "https://compute.googleapis.com/", Err: context.DeadlineExceeded,
+		}, provider.ReasonTimeout},
+		{"context ended", ended, context.Canceled, provider.ReasonTimeout},
+		{"connection refused", context.Background(), &url.Error{
+			Op: "Get", URL: "https://compute.googleapis.com/", Err: syscall.ECONNREFUSED,
+		}, provider.ReasonUnreachable},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			p := newTestProvider(&fakeAPI{probeErr: tc.err})
+
+			got := p.Check(tc.ctx)
+
+			if len(got) != 1 {
+				t.Fatalf("Check() returned %d targets, want 1", len(got))
+			}
+			if got[0].Ready || got[0].Reason != tc.want || got[0].Message != tc.err.Error() {
+				t.Errorf("Check() = %+v, want not ready, reason %s, message %q", got[0], tc.want, tc.err.Error())
+			}
+		})
+	}
+}
+
+// TestRealInstances_Probe runs the real SDK client against a local server,
+// since the point of Probe is how many requests the SDK makes.
+func TestRealInstances_Probe(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name     string
+		status   int
+		body     string
+		wantCode int
+	}{
+		{
+			name:   "fetches only the first page",
+			status: http.StatusOK,
+			body:   `{"items":{"zones/europe-west1-b":{"instances":[{"name":"proxy-a"}]}},"nextPageToken":"page-2"}`,
+		},
+		{
+			name:   "empty project",
+			status: http.StatusOK,
+			body:   `{}`,
+		},
+		{
+			name:     "api error keeps its status",
+			status:   http.StatusForbidden,
+			body:     `{"error":{"code":403,"message":"Compute Engine API has not been used","errors":[{"reason":"accessNotConfigured"}]}}`,
+			wantCode: http.StatusForbidden,
+		},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			var (
+				mu   sync.Mutex
+				reqs []*url.URL
+			)
+			srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+				mu.Lock()
+				reqs = append(reqs, r.URL)
+				mu.Unlock()
+				w.Header().Set("Content-Type", "application/json")
+				w.WriteHeader(tc.status)
+				_, _ = w.Write([]byte(tc.body))
+			}))
+			t.Cleanup(srv.Close)
+			ctx := context.Background()
+			c, err := compute.NewInstancesRESTClient(ctx, option.WithEndpoint(srv.URL), option.WithoutAuthentication())
+			if err != nil {
+				t.Fatalf("NewInstancesRESTClient: %v", err)
+			}
+			t.Cleanup(func() { _ = c.Close() })
+
+			err = (&realInstances{client: c}).Probe(ctx, &computepb.AggregatedListInstancesRequest{
+				Project:    "my-project",
+				MaxResults: proto.Uint32(1),
+			})
+
+			var gerr *googleapi.Error
+			switch {
+			case tc.wantCode == 0 && err != nil:
+				t.Fatalf("Probe() = %v, want nil", err)
+			case tc.wantCode != 0 && (!errors.As(err, &gerr) || gerr.Code != tc.wantCode):
+				t.Fatalf("Probe() = %v, want a googleapi error with code %d", err, tc.wantCode)
+			}
+			mu.Lock()
+			defer mu.Unlock()
+			if len(reqs) != 1 {
+				t.Fatalf("server saw %d requests, want 1", len(reqs))
+			}
+			if got := reqs[0].Query().Get("maxResults"); got != "1" {
+				t.Errorf("maxResults = %q, want 1", got)
+			}
+			if want := "/compute/v1/projects/my-project/aggregated/instances"; reqs[0].Path != want {
+				t.Errorf("path = %q, want %q", reqs[0].Path, want)
+			}
+		})
+	}
+}

internal/provider/gcp/gcp.go

@@ -33,6 +33,8 @@ type instancesAPI interface {
 	Get(ctx context.Context, req *computepb.GetInstanceRequest) (*computepb.Instance, error)
 	Delete(ctx context.Context, req *computepb.DeleteInstanceRequest) (opName string, err error)
 	AggregatedList(ctx context.Context, req *computepb.AggregatedListInstancesRequest) ([]*computepb.Instance, error)
+	// Probe fetches only the first page of an aggregated list; see Check.
+	Probe(ctx context.Context, req *computepb.AggregatedListInstancesRequest) error
 }
 
 // realInstances adapts *compute.InstancesClient to the seam.
@@ -221,17 +223,16 @@ func (p *Provider) Delete(ctx context.Context, providerID string) error {
 // entire GC sweep.
 func (p *Provider) ListByTag(ctx context.Context) ([]provider.Instance, error) {
 	log := p.logger(ctx)
-	filter := fmt.Sprintf("labels.%s = %s", provider.LabelManaged, provider.LabelManagedYes)
 	instances, err := p.api.AggregatedList(ctx, &computepb.AggregatedListInstancesRequest{
 		Project:              p.cfg.Project,
-		Filter:               proto.String(filter),
+		Filter:               proto.String(managedFilter),
 		ReturnPartialSuccess: proto.Bool(true),
 	})
 	if err != nil {
 		logAPIError(log, "list", err)
 		return nil, p.wrapErr("list", "", err)
 	}
-	log.V(1).Info("GCP instances listed", "filter", filter, "count", len(instances))
+	log.V(1).Info("GCP instances listed", "filter", managedFilter, "count", len(instances))
 	out := make([]provider.Instance, 0, len(instances))
 	for _, inst := range instances {
 		conv := toInstance(inst, lastPathSegment(inst.GetZone()))

internal/provider/gcp/gcp_test.go

@@ -31,6 +31,14 @@ type fakeAPI struct {
 	listReq   *computepb.AggregatedListInstancesRequest
 	listInsts []*computepb.Instance
 	listErr   error
+
+	probeReq *computepb.AggregatedListInstancesRequest
+	probeErr error
+}
+
+func (f *fakeAPI) Probe(_ context.Context, req *computepb.AggregatedListInstancesRequest) error {
+	f.probeReq = req
+	return f.probeErr
 }
 
 func (f *fakeAPI) Insert(_ context.Context, req *computepb.InsertInstanceRequest) (string, error) {

internal/provider/kubernetes/check.go

@@ -0,0 +1,53 @@
+package kubernetes
+
+import (
+	"context"
+	"errors"
+	"time"
+
+	corev1 "k8s.io/api/core/v1"
+	apierrors "k8s.io/apimachinery/pkg/api/errors"
+	"sigs.k8s.io/controller-runtime/pkg/client"
+
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+var _ provider.Checker = (*Provider)(nil)
+
+// Check lists at most one managed Pod across the cluster: the same client,
+// verb and scope ListByTag and orphan GC rely on, so a pass means the API
+// server answers and the operator's RBAC still covers them. It does not
+// prove create or delete rights.
+func (p *Provider) Check(ctx context.Context) []provider.TargetStatus {
+	start := time.Now()
+	var pods corev1.PodList
+	err := p.client.List(ctx, &pods, client.MatchingLabels{
+		provider.LabelManaged: provider.LabelManagedYes,
+	}, client.Limit(1))
+	ts := provider.TargetStatus{Name: "apiserver", Address: p.apiHost, Latency: time.Since(start)}
+	if err != nil {
+		ts.Reason = checkReason(ctx, err)
+		ts.Message = err.Error()
+		return []provider.TargetStatus{ts}
+	}
+	ts.Ready = true
+	ts.Reason = provider.ReasonReachable
+	return []provider.TargetStatus{ts}
+}
+
+func checkReason(ctx context.Context, err error) string {
+	var status apierrors.APIStatus
+	switch {
+	case ctx.Err() != nil, errors.Is(err, context.DeadlineExceeded),
+		apierrors.IsTimeout(err), apierrors.IsServerTimeout(err):
+		return provider.ReasonTimeout
+	case apierrors.IsUnauthorized(err):
+		return provider.ReasonUnauthorized
+	case apierrors.IsForbidden(err):
+		return provider.ReasonForbidden
+	case errors.As(err, &status):
+		return provider.ReasonAPIError
+	default:
+		return provider.ReasonUnreachable
+	}
+}

internal/provider/kubernetes/check_test.go

@@ -0,0 +1,114 @@
+package kubernetes
+
+import (
+	"context"
+	"errors"
+	"net/url"
+	"syscall"
+	"testing"
+
+	corev1 "k8s.io/api/core/v1"
+	apierrors "k8s.io/apimachinery/pkg/api/errors"
+	"k8s.io/apimachinery/pkg/runtime"
+	"k8s.io/apimachinery/pkg/runtime/schema"
+	"sigs.k8s.io/controller-runtime/pkg/client"
+	"sigs.k8s.io/controller-runtime/pkg/client/fake"
+	"sigs.k8s.io/controller-runtime/pkg/client/interceptor"
+
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+type listFunc func(ctx context.Context, c client.WithWatch, list client.ObjectList, opts ...client.ListOption) error
+
+func newCheckProvider(list listFunc) *Provider {
+	scheme := runtime.NewScheme()
+	if err := corev1.AddToScheme(scheme); err != nil {
+		panic(err)
+	}
+	c := fake.NewClientBuilder().WithScheme(scheme).WithInterceptorFuncs(interceptor.Funcs{List: list}).Build()
+	p := newWithClient(c, provider.ProviderConfig{Name: "kubernetes"})
+	p.apiHost = "https://10.96.0.1:443"
+	return p
+}
+
+func TestCheck_reachable(t *testing.T) {
+	t.Parallel()
+	p := newTestProvider(testPod("ns", "a", "uid-a", corev1.PodRunning, "10.0.0.1"))
+	p.apiHost = "https://10.96.0.1:443"
+
+	got := p.Check(context.Background())
+
+	if len(got) != 1 {
+		t.Fatalf("Check() returned %d targets, want 1", len(got))
+	}
+	ts := got[0]
+	if !ts.Ready || ts.Reason != provider.ReasonReachable || ts.Message != "" {
+		t.Errorf("Check() = %+v, want ready and Reachable", ts)
+	}
+	if ts.Name != "apiserver" || ts.Address != "https://10.96.0.1:443" {
+		t.Errorf("target = %q at %q, want apiserver at https://10.96.0.1:443", ts.Name, ts.Address)
+	}
+}
+
+func TestCheck_listOptions(t *testing.T) {
+	t.Parallel()
+	var got client.ListOptions
+	p := newCheckProvider(func(ctx context.Context, c client.WithWatch, list client.ObjectList, opts ...client.ListOption) error {
+		got.ApplyOptions(opts)
+		return c.List(ctx, list, opts...)
+	})
+
+	p.Check(context.Background())
+
+	if got.Limit != 1 {
+		t.Errorf("Limit = %d, want 1", got.Limit)
+	}
+	if got.Namespace != "" {
+		t.Errorf("Namespace = %q, want cluster-wide", got.Namespace)
+	}
+	want := provider.LabelManaged + "=" + provider.LabelManagedYes
+	if got.LabelSelector == nil || got.LabelSelector.String() != want {
+		t.Errorf("LabelSelector = %v, want %s", got.LabelSelector, want)
+	}
+}
+
+func TestCheck_failureReasons(t *testing.T) {
+	t.Parallel()
+	pods := schema.GroupResource{Resource: "pods"}
+	ended, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	tests := []struct {
+		name string
+		ctx  context.Context
+		err  error
+		want string
+	}{
+		{"forbidden", context.Background(), apierrors.NewForbidden(pods, "", errors.New("rbac denied")), provider.ReasonForbidden},
+		{"unauthorized", context.Background(), apierrors.NewUnauthorized("token expired"), provider.ReasonUnauthorized},
+		{"api timeout", context.Background(), apierrors.NewTimeoutError("slow etcd", 1), provider.ReasonTimeout},
+		{"server timeout", context.Background(), apierrors.NewServerTimeout(pods, "list", 1), provider.ReasonTimeout},
+		{"context ended", ended, context.Canceled, provider.ReasonTimeout},
+		{"internal error", context.Background(), apierrors.NewInternalError(errors.New("etcd down")), provider.ReasonAPIError},
+		{"connection refused", context.Background(), &url.Error{
+			Op: "Get", URL: "https://10.96.0.1:443/api/v1/pods", Err: syscall.ECONNREFUSED,
+		}, provider.ReasonUnreachable},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			p := newCheckProvider(func(context.Context, client.WithWatch, client.ObjectList, ...client.ListOption) error {
+				return tc.err
+			})
+
+			got := p.Check(tc.ctx)
+
+			if len(got) != 1 {
+				t.Fatalf("Check() returned %d targets, want 1", len(got))
+			}
+			if got[0].Ready || got[0].Reason != tc.want || got[0].Message != tc.err.Error() {
+				t.Errorf("Check() = %+v, want not ready, reason %s, message %q", got[0], tc.want, tc.err.Error())
+			}
+		})
+	}
+}

internal/provider/kubernetes/kubernetes.go

@@ -36,6 +36,8 @@ const defaultImage = "ubuntu/squid:6.6-24.04_edge"
 type Provider struct {
 	client client.Client
 	image  string
+	// apiHost is the API server URL, shown by Check; empty in tests.
+	apiHost string
 }
 
 // New builds a kubernetes Provider from its config block. Satisfies
@@ -64,7 +66,9 @@ func NewWithTransportWrapper(_ context.Context, cfg provider.ProviderConfig, wra
 	if err != nil {
 		return nil, fmt.Errorf("kubernetes provider %q: building client: %w", cfg.Name, err)
 	}
-	return newWithClient(c, cfg), nil
+	p := newWithClient(c, cfg)
+	p.apiHost = restCfg.Host
+	return p, nil
 }
 
 // newWithClient builds a Provider around an already-constructed client,

internal/provider/kvm/check.go

@@ -0,0 +1,166 @@
+package kvm
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"sync"
+	"time"
+
+	libvirt "github.com/digitalocean/go-libvirt"
+
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+var _ provider.Checker = (*Provider)(nil)
+
+// Reasons only the KVM check can tell apart; the rest are provider-wide.
+const (
+	reasonSSHMaterialUnavailable = "SSHMaterialUnavailable"
+	reasonSSHHandshakeFailed     = "SSHHandshakeFailed"
+	reasonLibvirtConnectFailed   = "LibvirtConnectFailed"
+	reasonLibvirtError           = "LibvirtError"
+	reasonPoolNotFound           = "PoolNotFound"
+	reasonPoolNotActive          = "PoolNotActive"
+)
+
+// hostFacts is what a successful check learns about a host.
+type hostFacts struct {
+	version   uint64
+	hostname  string
+	poolState libvirt.StoragePoolState
+	capacity  uint64
+	available uint64
+}
+
+// Check verifies every configured host concurrently, in kvm.hosts order.
+// It goes through the same pooled connection Create and Get use, so the
+// SSH and libvirt handshakes only happen on the first check or after the
+// link dropped; then two cheap RPCs prove the daemon answers, and the
+// storage pool VMs are created in must exist and be running.
+func (p *Provider) Check(ctx context.Context) []provider.TargetStatus {
+	out := make([]provider.TargetStatus, len(p.cfg.Hosts))
+	var wg sync.WaitGroup
+	for i, h := range p.cfg.Hosts {
+		wg.Go(func() { out[i] = p.checkHost(ctx, h) })
+	}
+	wg.Wait()
+	return out
+}
+
+// checkHost goes through probe, so a check that runs out of time abandons
+// its RPCs rather than closing the connection the reconciles share. fn
+// returns the RPC errors untouched, since a transport error still drops the
+// link; a pool that exists but is stopped says nothing about the link, so
+// that verdict is reached afterwards. facts is read only after probe
+// returned nil, since an abandoned fn may still write it.
+func (p *Provider) checkHost(ctx context.Context, host provider.KVMHost) provider.TargetStatus {
+	ts := provider.TargetStatus{
+		Name:    host.Name,
+		Address: host.Address,
+		Details: map[string]string{"user": host.User, "pool": host.Pool},
+	}
+	start := time.Now()
+	var facts hostFacts
+	err := p.hosts.probe(ctx, p.logger(ctx), host, func(api libvirtAPI) error {
+		version, err := api.ConnectGetLibVersion()
+		if err != nil {
+			return err
+		}
+		hostname, err := api.ConnectGetHostname()
+		if err != nil {
+			return err
+		}
+		pool, err := api.StoragePoolLookupByName(host.Pool)
+		if err != nil {
+			return err
+		}
+		state, capacity, _, available, err := api.StoragePoolGetInfo(pool)
+		if err != nil {
+			return err
+		}
+		facts = hostFacts{
+			version:   version,
+			hostname:  hostname,
+			poolState: libvirt.StoragePoolState(state),
+			capacity:  capacity,
+			available: available,
+		}
+		return nil
+	})
+	ts.Latency = time.Since(start)
+	if err != nil {
+		ts.Reason = checkReason(err)
+		ts.Message = err.Error()
+		return ts
+	}
+	ts.Details["hostname"] = facts.hostname
+	ts.Details["libvirtVersion"] = formatLibVersion(facts.version)
+	ts.Details["poolState"] = poolStateName(facts.poolState)
+	ts.Details["poolCapacity"] = formatGiB(facts.capacity)
+	ts.Details["poolAvailable"] = formatGiB(facts.available)
+	if facts.poolState != libvirt.StoragePoolRunning {
+		ts.Reason = reasonPoolNotActive
+		ts.Message = fmt.Sprintf("storage pool %q is %s", host.Pool, poolStateName(facts.poolState))
+		return ts
+	}
+	ts.Ready = true
+	ts.Reason = provider.ReasonReachable
+	return ts
+}
+
+// checkReason names where a failed check broke. The dial stage wins over
+// a timeout: "the SSH handshake did not finish" says more than "the check
+// ran out of time", and the message still carries the deadline.
+func checkReason(err error) string {
+	var de *dialError
+	if errors.As(err, &de) {
+		switch de.stage {
+		case stageSSHMaterial:
+			return reasonSSHMaterialUnavailable
+		case stageSSHHandshake:
+			return reasonSSHHandshakeFailed
+		case stageLibvirtConnect:
+			return reasonLibvirtConnectFailed
+		default:
+			return provider.ReasonUnreachable
+		}
+	}
+	switch {
+	case errors.Is(err, context.DeadlineExceeded), errors.Is(err, context.Canceled), errors.Is(err, errProbeBusy):
+		return provider.ReasonTimeout
+	case hasCode(err, libvirt.ErrNoStoragePool):
+		return reasonPoolNotFound
+	case isLibvirtError(err):
+		return reasonLibvirtError
+	default:
+		return provider.ReasonUnreachable
+	}
+}
+
+// formatLibVersion decodes libvirt's major*1e6 + minor*1e3 + micro.
+func formatLibVersion(v uint64) string {
+	return fmt.Sprintf("%d.%d.%d", v/1_000_000, v/1_000%1_000, v%1_000)
+}
+
+// poolStateName uses virsh's words for a pool state.
+func poolStateName(s libvirt.StoragePoolState) string {
+	switch s {
+	case libvirt.StoragePoolInactive:
+		return "inactive"
+	case libvirt.StoragePoolBuilding:
+		return "building"
+	case libvirt.StoragePoolRunning:
+		return "running"
+	case libvirt.StoragePoolDegraded:
+		return "degraded"
+	case libvirt.StoragePoolInaccessible:
+		return "inaccessible"
+	default:
+		return fmt.Sprintf("state %d", s)
+	}
+}
+
+func formatGiB(bytes uint64) string {
+	return fmt.Sprintf("%.1fGiB", float64(bytes)/(1<<30))
+}

internal/provider/kvm/check_test.go

@@ -0,0 +1,319 @@
+package kvm
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"maps"
+	"strings"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	libvirt "github.com/digitalocean/go-libvirt"
+
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+// withoutLatency zeroes the one field a test cannot predict.
+func withoutLatency(in []provider.TargetStatus) []provider.TargetStatus {
+	out := make([]provider.TargetStatus, len(in))
+	for i, ts := range in {
+		ts.Latency = 0
+		out[i] = ts
+	}
+	return out
+}
+
+func TestCheck_reachable(t *testing.T) {
+	t.Parallel()
+	api := newFakeAPI("default")
+	api.poolCapacity = 10 << 30
+	api.pools["default"]["disk.qcow2"] = &fakeVol{capacity: 4 << 30}
+	p := newTestProvider(hostFakes{"nas1": api}, "nas1")
+
+	got := withoutLatency(p.Check(context.Background()))
+
+	want := provider.TargetStatus{
+		Name:    "nas1",
+		Address: "nas1:22",
+		Ready:   true,
+		Reason:  provider.ReasonReachable,
+		Details: map[string]string{
+			"user":           "virt",
+			"pool":           "default",
+			"hostname":       "fake-host",
+			"libvirtVersion": "10.2.0",
+			"poolState":      "running",
+			"poolCapacity":   "10.0GiB",
+			"poolAvailable":  "6.0GiB",
+		},
+	}
+	if len(got) != 1 {
+		t.Fatalf("Check() returned %d targets, want 1", len(got))
+	}
+	if got[0].Name != want.Name || got[0].Address != want.Address || got[0].Ready != want.Ready ||
+		got[0].Reason != want.Reason || got[0].Message != "" || !maps.Equal(got[0].Details, want.Details) {
+		t.Errorf("Check() = %+v, want %+v", got[0], want)
+	}
+}
+
+func TestCheck_perHostVerdicts(t *testing.T) {
+	t.Parallel()
+	p := newTestProvider(hostFakes{"nas1": newFakeAPI("default"), "nas3": newFakeAPI("default")}, "nas1", "nas2", "nas3")
+
+	got := p.Check(context.Background())
+
+	if len(got) != 3 {
+		t.Fatalf("Check() returned %d targets, want 3", len(got))
+	}
+	for i, want := range []struct {
+		name  string
+		ready bool
+	}{{"nas1", true}, {"nas2", false}, {"nas3", true}} {
+		if got[i].Name != want.name || got[i].Ready != want.ready {
+			t.Errorf("target %d = %s ready=%v, want %s ready=%v", i, got[i].Name, got[i].Ready, want.name, want.ready)
+		}
+	}
+	if got[1].Reason != provider.ReasonUnreachable || !strings.Contains(got[1].Message, errHostOff.Error()) {
+		t.Errorf("unreachable host = %s %q, want Unreachable carrying %q", got[1].Reason, got[1].Message, errHostOff)
+	}
+	if got[1].Details["pool"] != "default" || got[1].Details["hostname"] != "" {
+		t.Errorf("unreachable host details = %v, want the configured pool and nothing learned", got[1].Details)
+	}
+}
+
+// TestCheck_libvirtVerdicts also checks twice per case: none of these
+// verdicts is evidence against the link, so the connection must survive.
+func TestCheck_libvirtVerdicts(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name        string
+		setup       func(api *fakeAPI)
+		pools       []string
+		wantReason  string
+		wantMessage string
+	}{
+		{
+			name:        "pool inactive",
+			pools:       []string{"default"},
+			setup:       func(api *fakeAPI) { api.poolStates["default"] = libvirt.StoragePoolInactive },
+			wantReason:  reasonPoolNotActive,
+			wantMessage: `storage pool "default" is inactive`,
+		},
+		{
+			name:        "pool missing",
+			pools:       []string{"other"},
+			wantReason:  reasonPoolNotFound,
+			wantMessage: "no storage pool with matching name 'default'",
+		},
+		{
+			name:        "daemon error",
+			pools:       []string{"default"},
+			setup:       func(api *fakeAPI) { api.fail["ConnectGetHostname"] = lvError(libvirt.ErrInternalError, "boom") },
+			wantReason:  reasonLibvirtError,
+			wantMessage: "boom",
+		},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			api := newFakeAPI(tc.pools...)
+			if tc.setup != nil {
+				tc.setup(api)
+			}
+			var dials atomic.Int32
+			dial := func(ctx context.Context, h provider.KVMHost) (*hostClient, error) {
+				dials.Add(1)
+				return hostFakes{"nas1": api}.dial(ctx, h)
+			}
+			pc := testConfig("nas1")
+			p := newWithDialer(pc.Name, *pc.KVM, dial, nil)
+
+			for range 2 {
+				got := p.Check(context.Background())
+				if got[0].Ready || got[0].Reason != tc.wantReason || !strings.Contains(got[0].Message, tc.wantMessage) {
+					t.Errorf("Check() = %+v, want not ready, %s, message containing %q", got[0], tc.wantReason, tc.wantMessage)
+				}
+			}
+			if n := dials.Load(); n != 1 {
+				t.Errorf("dialed %d times over two checks, want 1", n)
+			}
+		})
+	}
+}
+
+func TestCheck_hostsCheckedConcurrently(t *testing.T) {
+	t.Parallel()
+	nas2Dialing := make(chan struct{})
+	dial := func(ctx context.Context, h provider.KVMHost) (*hostClient, error) {
+		switch h.Name {
+		case "nas1":
+			select {
+			case <-nas2Dialing:
+			case <-ctx.Done():
+				return nil, ctx.Err()
+			}
+		case "nas2":
+			close(nas2Dialing)
+		}
+		return &hostClient{api: newFakeAPI("default"), close: func() {}}, nil
+	}
+	pc := testConfig("nas1", "nas2")
+	p := newWithDialer(pc.Name, *pc.KVM, dial, nil)
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	got := p.Check(ctx)
+
+	for _, ts := range got {
+		if !ts.Ready {
+			t.Errorf("%s = %s %q, want ready: a sequential check would time out nas1", ts.Name, ts.Reason, ts.Message)
+		}
+	}
+}
+
+// A daemon that stopped answering times the check out, but the connection
+// the reconciles share stays up, and the next check does not stack a
+// second hung RPC on top of the first.
+func TestCheck_timeoutKeepsTheConnection(t *testing.T) {
+	t.Parallel()
+	api := newFakeAPI("default")
+	block := make(chan struct{})
+	t.Cleanup(func() { close(block) })
+	blocked := &blockingAPI{fakeAPI: api, block: block}
+	var closed atomic.Bool
+	dial := func(context.Context, provider.KVMHost) (*hostClient, error) {
+		return &hostClient{api: blocked, close: func() { closed.Store(true) }}, nil
+	}
+	pc := testConfig("nas1")
+	p := newWithDialer(pc.Name, *pc.KVM, dial, nil)
+	ctx, cancel := context.WithTimeout(context.Background(), 50*time.Millisecond)
+	defer cancel()
+
+	got := p.Check(ctx)
+
+	if got[0].Ready || got[0].Reason != provider.ReasonTimeout {
+		t.Errorf("Check() = %+v, want Timeout", got[0])
+	}
+	if closed.Load() {
+		t.Error("the timed-out check closed the connection the reconciles share")
+	}
+	again := p.Check(context.Background())
+	if again[0].Reason != provider.ReasonTimeout || !strings.Contains(again[0].Message, "still waiting") {
+		t.Errorf("second Check() = %+v, want an immediate Timeout while the first still waits", again[0])
+	}
+}
+
+// A reconcile redialing a dead host holds the host's slot for minutes; the
+// check must give up at its own deadline rather than wait that out.
+func TestCheck_doesNotWaitBehindAReconcileDial(t *testing.T) {
+	t.Parallel()
+	release := make(chan struct{})
+	dialing := make(chan struct{})
+	var once sync.Once
+	dial := func(context.Context, provider.KVMHost) (*hostClient, error) {
+		once.Do(func() { close(dialing) })
+		<-release
+		return nil, errHostOff
+	}
+	pc := testConfig("nas1")
+	p := newWithDialer(pc.Name, *pc.KVM, dial, nil)
+	reconcile := make(chan error, 1)
+	go func() {
+		_, err := p.Get(context.Background(), formatProviderID("nas1", testName))
+		reconcile <- err
+	}()
+	<-dialing
+	ctx, cancel := context.WithTimeout(context.Background(), 50*time.Millisecond)
+	defer cancel()
+	start := time.Now()
+
+	got := p.Check(ctx)
+
+	if waited := time.Since(start); waited > time.Second {
+		t.Errorf("Check() took %s behind the reconcile's dial, want about 50ms", waited)
+	}
+	if got[0].Reason != provider.ReasonTimeout {
+		t.Errorf("Check() = %+v, want Timeout", got[0])
+	}
+	close(release)
+	<-reconcile
+}
+
+// blockingAPI hangs in ConnectGetLibVersion until block closes, like a
+// daemon that stopped answering.
+type blockingAPI struct {
+	*fakeAPI
+	block chan struct{}
+}
+
+func (b *blockingAPI) ConnectGetLibVersion() (uint64, error) {
+	<-b.block
+	return 0, io.EOF
+}
+
+func TestCheckReason(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name string
+		err  error
+		want string
+	}{
+		{"ssh material", &dialError{stage: stageSSHMaterial, err: fmt.Errorf("%w: parsing key", provider.ErrPermanent)}, reasonSSHMaterialUnavailable},
+		{"tcp", &dialError{stage: stageTCP, err: errors.New("ssh dial: connection refused")}, provider.ReasonUnreachable},
+		{"ssh handshake", &dialError{stage: stageSSHHandshake, err: errors.New("ssh handshake: unable to authenticate")}, reasonSSHHandshakeFailed},
+		{"libvirt connect", &dialError{stage: stageLibvirtConnect, err: errors.New("libvirt connect via /sock: EOF")}, reasonLibvirtConnectFailed},
+		{"stage survives the pool's wrapping", fmt.Errorf("connecting to host nas1 (nas1:22): %w",
+			&dialError{stage: stageSSHHandshake, err: context.DeadlineExceeded}), reasonSSHHandshakeFailed},
+		{"deadline", context.DeadlineExceeded, provider.ReasonTimeout},
+		{"cancelled", context.Canceled, provider.ReasonTimeout},
+		{"previous check still waiting", errProbeBusy, provider.ReasonTimeout},
+		{"no pool", lvError(libvirt.ErrNoStoragePool, "no pool"), reasonPoolNotFound},
+		{"other libvirt error", lvError(libvirt.ErrInternalError, "boom"), reasonLibvirtError},
+		{"link died mid-call", io.EOF, provider.ReasonUnreachable},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			if got := checkReason(tc.err); got != tc.want {
+				t.Errorf("checkReason(%v) = %s, want %s", tc.err, got, tc.want)
+			}
+		})
+	}
+}
+
+func TestDialError_keepsClassification(t *testing.T) {
+	t.Parallel()
+	err := fmt.Errorf("connecting to host nas1: %w",
+		&dialError{stage: stageSSHMaterial, err: fmt.Errorf("%w: parsing kvm.sshKeyFile: bad", provider.ErrPermanent)})
+
+	if got := classify(err); !errors.Is(got, provider.ErrPermanent) {
+		t.Errorf("classify() = %v, want ErrPermanent through the dialError", got)
+	}
+	if !isClassified(err) {
+		t.Error("isClassified() = false, want true")
+	}
+	if want := "connecting to host nas1: permanent: parsing kvm.sshKeyFile: bad"; !strings.HasSuffix(err.Error(), "parsing kvm.sshKeyFile: bad") {
+		t.Errorf("message = %q, want it unchanged, like %q", err.Error(), want)
+	}
+}
+
+func TestFormatLibVersion(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		in   uint64
+		want string
+	}{
+		{10002000, "10.2.0"},
+		{11006001, "11.6.1"},
+		{9000000, "9.0.0"},
+	}
+	for _, tc := range tests {
+		if got := formatLibVersion(tc.in); got != tc.want {
+			t.Errorf("formatLibVersion(%d) = %s, want %s", tc.in, got, tc.want)
+		}
+	}
+}

internal/provider/kvm/conn.go

@@ -5,9 +5,11 @@ import (
 	"errors"
 	"fmt"
 	"io"
+	"io/fs"
 	"net"
 	"os"
 	"sync"
+	"sync/atomic"
 	"time"
 
 	libvirt "github.com/digitalocean/go-libvirt"
@@ -43,6 +45,9 @@ type libvirtAPI interface {
 	StorageVolUpload(vol libvirt.StorageVol, r io.Reader, offset, length uint64, flags libvirt.StorageVolUploadFlags) error
 	StorageVolDownload(vol libvirt.StorageVol, w io.Writer, offset, length uint64, flags libvirt.StorageVolDownloadFlags) error
 	StorageVolDelete(vol libvirt.StorageVol, flags libvirt.StorageVolDeleteFlags) error
+	ConnectGetLibVersion() (uint64, error)
+	ConnectGetHostname() (string, error)
+	StoragePoolGetInfo(pool libvirt.StoragePool) (state uint8, capacity, allocation, available uint64, err error)
 }
 
 // hostClient is one live connection to a hypervisor.
@@ -67,6 +72,27 @@ func (c *hostClient) alive() bool {
 // real one; tests inject their own.
 type hostDialer func(ctx context.Context, host provider.KVMHost) (*hostClient, error)
 
+// dialStage names the step of a connection attempt that failed, so the
+// health check can say why a host is unreachable without parsing messages.
+type dialStage string
+
+const (
+	stageSSHMaterial    dialStage = "ssh-material"
+	stageTCP            dialStage = "tcp"
+	stageSSHHandshake   dialStage = "ssh-handshake"
+	stageLibvirtConnect dialStage = "libvirt-connect"
+)
+
+// dialError tags a dial failure with its stage. It keeps the message and
+// unwraps to the cause, so classify sees exactly what it saw before.
+type dialError struct {
+	stage dialStage
+	err   error
+}
+
+func (e *dialError) Error() string { return e.err.Error() }
+func (e *dialError) Unwrap() error { return e.err }
+
 // hostPool caches one connection per host, dialing lazily and redialing
 // after the connection drops. Dials for different hosts don't block each
 // other; concurrent callers for the same host share a single dial.
@@ -77,9 +103,16 @@ type hostPool struct {
 	entries map[string]*hostEntry
 }
 
+// hostEntry is one host's slot in the pool. lock is a one-slot semaphore
+// rather than a mutex so that waiting for it can end with the caller's
+// context: whoever holds it may be dialing a dead host for as long as the
+// dial takes.
 type hostEntry struct {
-	mu     sync.Mutex
+	lock   chan struct{}
 	client *hostClient
+	// probing is set while a probe's fn runs, including one whose caller
+	// has already given up on it.
+	probing atomic.Bool
 }
 
 func newHostPool(dial hostDialer) *hostPool {
@@ -91,16 +124,37 @@ func (p *hostPool) entry(name string) *hostEntry {
 	defer p.mu.Unlock()
 	e, ok := p.entries[name]
 	if !ok {
-		e = &hostEntry{}
+		e = &hostEntry{lock: make(chan struct{}, 1)}
 		p.entries[name] = e
 	}
 	return e
 }
 
+// acquire takes e.lock. Only waiting is bounded by ctx: a free slot is
+// taken even when ctx has already ended, so an uncontended caller behaves
+// as it did with a mutex, and a contended one gives up with ctx's error.
+func (e *hostEntry) acquire(ctx context.Context) error {
+	select {
+	case e.lock <- struct{}{}:
+		return nil
+	default:
+	}
+	select {
+	case e.lock <- struct{}{}:
+		return nil
+	case <-ctx.Done():
+		return ctx.Err()
+	}
+}
+
+func (e *hostEntry) release() { <-e.lock }
+
 func (p *hostPool) get(ctx context.Context, log logr.Logger, host provider.KVMHost) (*hostClient, error) {
 	e := p.entry(host.Name)
-	e.mu.Lock()
-	defer e.mu.Unlock()
+	if err := e.acquire(ctx); err != nil {
+		return nil, fmt.Errorf("waiting for the connection to host %s (%s): %w", host.Name, host.Address, err)
+	}
+	defer e.release()
 	if e.client != nil {
 		if e.client.alive() {
 			return e.client, nil
@@ -124,8 +178,8 @@ func (p *hostPool) get(ctx context.Context, log logr.Logger, host provider.KVMHo
 // connection that was already replaced by a redial is left alone.
 func (p *hostPool) drop(host string, c *hostClient) {
 	e := p.entry(host)
-	e.mu.Lock()
-	defer e.mu.Unlock()
+	e.lock <- struct{}{}
+	defer e.release()
 	if e.client == c {
 		e.client = nil
 	}
@@ -135,29 +189,74 @@ func (p *hostPool) drop(host string, c *hostClient) {
 // call runs fn against host's connection. go-libvirt RPCs take no context
 // and block until the daemon answers, so when ctx ends first the connection
 // is closed — which unblocks fn with an error — and ctx.Err() is returned.
-// A transport error (anything that is not a libvirt.Error) also drops the
-// cached connection so the next call redials instead of reusing a dead
-// link. When log has V(1) enabled, every RPC fn makes is traced through it
-// (see tracingAPI), as are connection events.
+// A transport error also drops the cached connection so the next call
+// redials instead of reusing a dead link. "Transport error" means anything
+// that is neither a libvirt.Error nor an error fn already classified with a
+// provider sentinel (a failed seed build, a bad placement): those are local
+// verdicts, not evidence about the link. When log has V(1) enabled, every
+// RPC fn makes is traced through it (see tracingAPI), as are connection
+// events.
 func (p *hostPool) call(ctx context.Context, log logr.Logger, host provider.KVMHost, fn func(api libvirtAPI) error) error {
 	log = log.WithName("libvirt")
 	c, err := p.get(ctx, log, host)
 	if err != nil {
 		return err
 	}
-	api := c.api
-	if log.V(1).Enabled() {
-		api = &tracingAPI{inner: c.api, log: log}
-	}
-	err = awaitOrCancel(ctx, func() error { return fn(api) }, func() {
+	err = awaitOrCancel(ctx, func() error { return fn(traced(c, log)) }, func() {
 		log.V(1).Info("libvirt connection dropped", "reason", "context cancelled", "error", ctx.Err().Error())
 		p.drop(host.Name, c)
 	})
-	if err != nil && !isLibvirtError(err) && !errors.Is(err, ctx.Err()) {
+	p.dropOnTransportError(ctx, log, host, c, err)
+	return err
+}
+
+// errProbeBusy refuses a probe while the previous one on the same host has
+// not come back yet.
+var errProbeBusy = errors.New("the previous check of this host is still waiting for libvirt to answer")
+
+// probe is call for the health check, which must never cost the reconciles
+// their connection: when ctx ends first, fn is abandoned and the connection
+// left alone. A link that is really dead is closed by the keepalive, which
+// ends fn too. At most one probe per host runs at a time, an abandoned one
+// included, so a daemon that stopped answering cannot pile up goroutines;
+// until it answers, further probes get errProbeBusy.
+func (p *hostPool) probe(ctx context.Context, log logr.Logger, host provider.KVMHost, fn func(api libvirtAPI) error) error {
+	log = log.WithName("libvirt")
+	e := p.entry(host.Name)
+	if !e.probing.CompareAndSwap(false, true) {
+		return errProbeBusy
+	}
+	c, err := p.get(ctx, log, host)
+	if err != nil {
+		e.probing.Store(false)
+		return err
+	}
+	err = awaitOrCancel(ctx, func() error {
+		defer e.probing.Store(false)
+		return fn(traced(c, log))
+	}, func() {
+		log.V(1).Info("libvirt probe abandoned, connection kept", "error", ctx.Err().Error())
+	})
+	p.dropOnTransportError(ctx, log, host, c, err)
+	return err
+}
+
+// traced returns c's API, logging every RPC when log has V(1) enabled.
+func traced(c *hostClient, log logr.Logger) libvirtAPI {
+	if log.V(1).Enabled() {
+		return &tracingAPI{inner: c.api, log: log}
+	}
+	return c.api
+}
+
+// dropOnTransportError drops c when err is evidence against the link:
+// neither a libvirt error, nor a verdict already classified, nor the
+// caller's own context ending.
+func (p *hostPool) dropOnTransportError(ctx context.Context, log logr.Logger, host provider.KVMHost, c *hostClient, err error) {
+	if err != nil && !isLibvirtError(err) && !isClassified(err) && !errors.Is(err, ctx.Err()) {
 		log.V(1).Info("libvirt connection dropped", "reason", "transport error", "error", err.Error())
 		p.drop(host.Name, c)
 	}
-	return err
 }
 
 // awaitOrCancel runs fn in the background and returns its result, unless ctx
@@ -177,14 +276,17 @@ func awaitOrCancel(ctx context.Context, fn func() error, abort func()) error {
 }
 
 // sshAuth holds the pieces of the SSH client config shared by every host:
-// the private key and the host-key policy, both loaded once at startup so a
-// missing or unreadable file fails the operator fast rather than on the
-// first Create.
+// the private key and the host-key policy. Both are loaded on the first
+// connection rather than at startup — see lazySSH.
 type sshAuth struct {
 	signer  ssh.Signer
 	hostKey ssh.HostKeyCallback
 }
 
+// newSSHAuth loads the key material. A file that is not there yet stays
+// unclassified, so the provider treats it as transient and heals when the
+// Secret appears; content that will never parse is permanent, since
+// retrying cannot fix it.
 func newSSHAuth(cfg provider.KVMConfig) (*sshAuth, error) {
 	key, err := os.ReadFile(cfg.SSHKeyFile)
 	if err != nil {
@@ -192,23 +294,69 @@ func newSSHAuth(cfg provider.KVMConfig) (*sshAuth, error) {
 	}
 	signer, err := ssh.ParsePrivateKey(key)
 	if err != nil {
-		return nil, fmt.Errorf("parsing kvm.sshKeyFile: %w", err)
+		return nil, fmt.Errorf("%w: parsing kvm.sshKeyFile: %v", provider.ErrPermanent, err)
 	}
 	hostKey := ssh.InsecureIgnoreHostKey() //nolint:gosec // opt-in via config, documented as insecure
 	if !cfg.InsecureIgnoreHostKey {
 		hostKey, err = knownhosts.New(cfg.KnownHostsFile)
-		if err != nil {
+		if errors.Is(err, fs.ErrNotExist) {
 			return nil, fmt.Errorf("loading kvm.knownHostsFile: %w", err)
 		}
+		if err != nil {
+			return nil, fmt.Errorf("%w: loading kvm.knownHostsFile: %v", provider.ErrPermanent, err)
+		}
 	}
 	return &sshAuth{signer: signer, hostKey: hostKey}, nil
 }
 
-// dialer returns the production hostDialer: SSH to the host, forward its
-// libvirt unix socket through the connection, and open a libvirt session
-// over that.
-func (a *sshAuth) dialer() hostDialer {
-	return func(ctx context.Context, host provider.KVMHost) (*hostClient, error) {
+// lazySSH defers loading the SSH material until it is first needed and
+// caches it once it works. The kvm-ssh Secret is an optional mount: the
+// manager has to start without it and keep every other provider running,
+// and a Secret created afterwards has to heal the KVM proxies on the next
+// reconcile rather than needing a restart.
+type lazySSH struct {
+	cfg provider.KVMConfig
+
+	mu   sync.Mutex
+	auth *sshAuth
+}
+
+// newLazySSH keeps only the SSH fields of cfg, since the loader lives as
+// long as the Provider.
+func newLazySSH(cfg provider.KVMConfig) *lazySSH {
+	return &lazySSH{cfg: provider.KVMConfig{
+		SSHKeyFile:            cfg.SSHKeyFile,
+		KnownHostsFile:        cfg.KnownHostsFile,
+		InsecureIgnoreHostKey: cfg.InsecureIgnoreHostKey,
+	}}
+}
+
+func (l *lazySSH) load() (*sshAuth, error) {
+	l.mu.Lock()
+	defer l.mu.Unlock()
+	if l.auth != nil {
+		return l.auth, nil
+	}
+	auth, err := newSSHAuth(l.cfg)
+	if err != nil {
+		return nil, err
+	}
+	l.auth = auth
+	return auth, nil
+}
+
+// dial is the production hostDialer.
+func (l *lazySSH) dial(ctx context.Context, host provider.KVMHost) (*hostClient, error) {
+	a, err := l.load()
+	if err != nil {
+		return nil, &dialError{stage: stageSSHMaterial, err: err}
+	}
+	return a.dial(ctx, host)
+}
+
+// dial SSHes to the host, forwards its libvirt unix socket through the
+// connection, and opens a libvirt session over that.
+func (a *sshAuth) dial(ctx context.Context, host provider.KVMHost) (*hostClient, error) {
 	client, err := a.dialSSH(ctx, host)
 	if err != nil {
 		return nil, err
@@ -219,7 +367,6 @@ func (a *sshAuth) dialer() hostDialer {
 		return nil, err
 	}
 	return hc, nil
-	}
 }
 
 // dialSSH establishes the SSH connection. The handshake is bounded by ctx's
@@ -233,7 +380,7 @@ func (a *sshAuth) dialSSH(ctx context.Context, host provider.KVMHost) (*ssh.Clie
 	var d net.Dialer
 	nc, err := d.DialContext(ctx, "tcp", host.Address)
 	if err != nil {
-		return nil, fmt.Errorf("ssh dial: %w", err)
+		return nil, &dialError{stage: stageTCP, err: fmt.Errorf("ssh dial: %w", err)}
 	}
 	_ = nc.SetDeadline(deadline)
 	conn, chans, reqs, err := ssh.NewClientConn(nc, host.Address, &ssh.ClientConfig{
@@ -243,7 +390,7 @@ func (a *sshAuth) dialSSH(ctx context.Context, host provider.KVMHost) (*ssh.Clie
 	})
 	if err != nil {
 		nc.Close()
-		return nil, fmt.Errorf("ssh handshake: %w", err)
+		return nil, &dialError{stage: stageSSHHandshake, err: fmt.Errorf("ssh handshake: %w", err)}
 	}
 	_ = nc.SetDeadline(time.Time{})
 	return ssh.NewClient(conn, chans, reqs), nil
@@ -267,7 +414,7 @@ func (u unixForward) Dial() (net.Conn, error) {
 func connectLibvirt(ctx context.Context, client *ssh.Client, socket string) (*hostClient, error) {
 	lv := libvirt.NewWithDialer(unixForward{client: client, path: socket})
 	if err := awaitOrCancel(ctx, lv.Connect, func() { client.Close() }); err != nil {
-		return nil, fmt.Errorf("libvirt connect via %s: %w", socket, err)
+		return nil, &dialError{stage: stageLibvirtConnect, err: fmt.Errorf("libvirt connect via %s: %w", socket, err)}
 	}
 	stop := make(chan struct{})
 	var once sync.Once

internal/provider/kvm/conn_test.go

@@ -7,6 +7,7 @@ import (
 	"crypto/rand"
 	"encoding/pem"
 	"errors"
+	"fmt"
 	"io"
 	"net"
 	"os"
@@ -109,41 +110,48 @@ func TestHostPool_call_redialsAfterDisconnect(t *testing.T) {
 	}
 }
 
-func TestHostPool_call_transportErrorDropsConnection(t *testing.T) {
+// Only a transport error is evidence about the link. A daemon-reported
+// libvirt error, or a verdict fn reached locally, must not cost the
+// connection.
+func TestHostPool_call_dropsConnectionOnlyOnTransportError(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name        string
+		err         error
+		wantDropped bool
+	}{
+		{"transport error", io.ErrUnexpectedEOF, true},
+		{"libvirt error", lvErr(libvirt.ErrNoDomain), false},
+		{"locally classified error", fmt.Errorf("%w: creating iso writer: read-only file system", provider.ErrPermanent), false},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
 			t.Parallel()
 			d := &fakeDialer{}
 			p := newHostPool(d.dial)
-	err := p.call(context.Background(), discard, testHost, func(libvirtAPI) error { return io.ErrUnexpectedEOF })
-	if !errors.Is(err, io.ErrUnexpectedEOF) {
-		t.Fatalf("call() error = %v, want the transport error passed through", err)
+			log, snapshot := captureLogger(1)
+			err := p.call(context.Background(), log, testHost, func(libvirtAPI) error { return tc.err })
+			if !errors.Is(err, tc.err) {
+				t.Fatalf("call() error = %v, want %v passed through", err, tc.err)
 			}
-	if !d.clients[0].closed.Load() {
-		t.Error("connection not closed after a transport error")
+			if got := d.clients[0].closed.Load(); got != tc.wantDropped {
+				t.Errorf("connection closed = %v, want %v", got, tc.wantDropped)
+			}
+			joined := strings.Join(snapshot(), "\n")
+			if logged := strings.Contains(joined, `"msg"="libvirt connection dropped"`) && strings.Contains(joined, `"reason"="transport error"`); logged != tc.wantDropped {
+				t.Errorf("transport drop logged = %v, want %v:\n%s", logged, tc.wantDropped, joined)
 			}
 			if err := p.call(context.Background(), discard, testHost, noop); err != nil {
-		t.Fatalf("call() after transport error = %v", err)
+				t.Fatalf("next call() error = %v", err)
 			}
-	if d.dialCount() != 2 {
-		t.Errorf("dials = %d, want 2", d.dialCount())
+			wantDials := 1
+			if tc.wantDropped {
+				wantDials = 2
 			}
-}
-
-func TestHostPool_call_libvirtErrorKeepsConnection(t *testing.T) {
-	t.Parallel()
-	d := &fakeDialer{}
-	p := newHostPool(d.dial)
-	err := p.call(context.Background(), discard, testHost, func(libvirtAPI) error { return lvErr(libvirt.ErrNoDomain) })
-	if !hasCode(err, libvirt.ErrNoDomain) {
-		t.Fatalf("call() error = %v, want the libvirt error passed through", err)
+			if d.dialCount() != wantDials {
+				t.Errorf("dials = %d, want %d", d.dialCount(), wantDials)
 			}
-	if err := p.call(context.Background(), discard, testHost, noop); err != nil {
-		t.Fatalf("call() error = %v", err)
-	}
-	if d.dialCount() != 1 {
-		t.Errorf("dials = %d, want 1: a daemon-reported error must not cost the connection", d.dialCount())
-	}
-	if d.clients[0].closed.Load() {
-		t.Error("connection closed after a libvirt error")
+		})
 	}
 }
 
@@ -224,6 +232,97 @@ func TestHostPool_call_hostsAreIndependent(t *testing.T) {
 	}
 }
 
+// A dial to a dead host holds the host's slot for as long as it takes; a
+// caller with a deadline must not wait past it.
+func TestHostPool_get_waitingRespectsContext(t *testing.T) {
+	t.Parallel()
+	release := make(chan struct{})
+	dialing := make(chan struct{})
+	var once sync.Once
+	dial := func(context.Context, provider.KVMHost) (*hostClient, error) {
+		once.Do(func() { close(dialing) })
+		<-release
+		return nil, errors.New("ssh dial: connection timed out")
+	}
+	p := newHostPool(dial)
+	slow := make(chan error, 1)
+	go func() { slow <- p.call(context.Background(), discard, testHost, noop) }()
+	<-dialing
+	ctx, cancel := context.WithTimeout(context.Background(), 50*time.Millisecond)
+	defer cancel()
+	start := time.Now()
+
+	err := p.call(ctx, discard, testHost, noop)
+
+	if !errors.Is(err, context.DeadlineExceeded) {
+		t.Errorf("call() error = %v, want the caller's deadline", err)
+	}
+	if waited := time.Since(start); waited > time.Second {
+		t.Errorf("call() waited %s behind another caller's dial, want about 50ms", waited)
+	}
+	close(release)
+	if err := <-slow; err == nil {
+		t.Error("the blocked dial succeeded, want its error")
+	}
+}
+
+func TestHostPool_probe_abandonsWithoutClosing(t *testing.T) {
+	t.Parallel()
+	d := &fakeDialer{}
+	p := newHostPool(d.dial)
+	unblock := make(chan struct{})
+	ctx, cancel := context.WithTimeout(context.Background(), 20*time.Millisecond)
+	defer cancel()
+
+	err := p.probe(ctx, discard, testHost, func(libvirtAPI) error {
+		<-unblock // a daemon that stopped answering
+		return nil
+	})
+
+	if !errors.Is(err, context.DeadlineExceeded) {
+		t.Fatalf("probe() error = %v, want DeadlineExceeded", err)
+	}
+	if d.clients[0].closed.Load() {
+		t.Error("probe closed the shared connection when it gave up")
+	}
+	if err := p.probe(context.Background(), discard, testHost, noop); !errors.Is(err, errProbeBusy) {
+		t.Errorf("probe() while the abandoned one still waits = %v, want errProbeBusy", err)
+	}
+	if err := p.call(context.Background(), discard, testHost, noop); err != nil {
+		t.Errorf("call() on the kept connection = %v", err)
+	}
+	close(unblock)
+	deadline := time.Now().Add(time.Second)
+	for {
+		err := p.probe(context.Background(), discard, testHost, noop)
+		if err == nil {
+			break
+		}
+		if !errors.Is(err, errProbeBusy) || time.Now().After(deadline) {
+			t.Fatalf("probe() after the abandoned one returned = %v, want nil", err)
+		}
+		time.Sleep(time.Millisecond)
+	}
+	if d.dialCount() != 1 {
+		t.Errorf("dials = %d, want 1: nothing should have redialed", d.dialCount())
+	}
+}
+
+func TestHostPool_probe_dropsConnectionOnTransportError(t *testing.T) {
+	t.Parallel()
+	d := &fakeDialer{}
+	p := newHostPool(d.dial)
+
+	err := p.probe(context.Background(), discard, testHost, func(libvirtAPI) error { return io.ErrUnexpectedEOF })
+
+	if !errors.Is(err, io.ErrUnexpectedEOF) {
+		t.Fatalf("probe() error = %v, want io.ErrUnexpectedEOF passed through", err)
+	}
+	if !d.clients[0].closed.Load() {
+		t.Error("a transport error during a probe left the connection cached")
+	}
+}
+
 // --- SSH dial against an in-process server ---
 
 func newKeyPair(t *testing.T) (ssh.Signer, []byte) {
@@ -387,13 +486,19 @@ func newSSHFixture(t *testing.T, answerGlobalRequests bool) *sshFixture {
 	}
 }
 
-func TestSSHAuth_dialSSH_forwardsUnixSocket(t *testing.T) {
-	t.Parallel()
-	f := newSSHFixture(t, true)
-	auth, err := newSSHAuth(f.cfg)
+func mustSSHAuth(t *testing.T, cfg provider.KVMConfig) *sshAuth {
+	t.Helper()
+	auth, err := newSSHAuth(cfg)
 	if err != nil {
 		t.Fatalf("newSSHAuth() error = %v", err)
 	}
+	return auth
+}
+
+func TestSSHAuth_dialSSH_forwardsUnixSocket(t *testing.T) {
+	t.Parallel()
+	f := newSSHFixture(t, true)
+	auth := mustSSHAuth(t, f.cfg)
 	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
 	defer cancel()
 	client, err := auth.dialSSH(ctx, f.host)
@@ -427,10 +532,7 @@ func TestSSHAuth_dialSSH_rejectsUnknownHostKey(t *testing.T) {
 	if err := os.WriteFile(f.cfg.KnownHostsFile, []byte(line+"\n"), 0o600); err != nil {
 		t.Fatal(err)
 	}
-	auth, err := newSSHAuth(f.cfg)
-	if err != nil {
-		t.Fatalf("newSSHAuth() error = %v", err)
-	}
+	auth := mustSSHAuth(t, f.cfg)
 	client, err := auth.dialSSH(context.Background(), f.host)
 	if err == nil {
 		client.Close()
@@ -447,11 +549,7 @@ func TestSSHAuth_dialSSH_insecureIgnoreHostKey(t *testing.T) {
 	f := newSSHFixture(t, true)
 	f.cfg.KnownHostsFile = ""
 	f.cfg.InsecureIgnoreHostKey = true
-	auth, err := newSSHAuth(f.cfg)
-	if err != nil {
-		t.Fatalf("newSSHAuth() error = %v", err)
-	}
-	client, err := auth.dialSSH(context.Background(), f.host)
+	client, err := mustSSHAuth(t, f.cfg).dialSSH(context.Background(), f.host)
 	if err != nil {
 		t.Fatalf("dialSSH() error = %v", err)
 	}
@@ -465,11 +563,7 @@ func TestSSHAuth_dialSSH_rejectsWrongClientKey(t *testing.T) {
 	if err := os.WriteFile(f.keyFile, otherPEM, 0o600); err != nil {
 		t.Fatal(err)
 	}
-	auth, err := newSSHAuth(f.cfg)
-	if err != nil {
-		t.Fatalf("newSSHAuth() error = %v", err)
-	}
-	client, err := auth.dialSSH(context.Background(), f.host)
+	client, err := mustSSHAuth(t, f.cfg).dialSSH(context.Background(), f.host)
 	if err == nil {
 		client.Close()
 		t.Fatal("dialSSH() succeeded with a key the server does not know")
@@ -482,13 +576,9 @@ func TestSSHAuth_dialSSH_rejectsWrongClientKey(t *testing.T) {
 func TestSSHAuth_dialSSH_cancelledContext(t *testing.T) {
 	t.Parallel()
 	f := newSSHFixture(t, true)
-	auth, err := newSSHAuth(f.cfg)
-	if err != nil {
-		t.Fatalf("newSSHAuth() error = %v", err)
-	}
 	ctx, cancel := context.WithCancel(context.Background())
 	cancel()
-	if client, err := auth.dialSSH(ctx, f.host); err == nil {
+	if client, err := mustSSHAuth(t, f.cfg).dialSSH(ctx, f.host); err == nil {
 		client.Close()
 		t.Fatal("dialSSH() succeeded with an already-cancelled context")
 	}
@@ -524,14 +614,181 @@ func TestNewSSHAuth_missingFiles(t *testing.T) {
 	}
 }
 
+// The kvm-ssh Secret is an optional mount, so the provider has to be
+// constructible without it and has to start working once it appears — the
+// alternative was a constructor error that crash-looped the whole manager.
+func TestLazySSH_load_healsWhenTheKeyAppears(t *testing.T) {
+	t.Parallel()
+	f := newSSHFixture(t, true)
+	keyPEM, err := os.ReadFile(f.keyFile)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if err := os.Remove(f.keyFile); err != nil {
+		t.Fatal(err)
+	}
+	auth := newLazySSH(f.cfg).load
+
+	if _, err := auth(); err == nil {
+		t.Fatal("loading with no key file succeeded, want an error")
+	} else if class := classify(err); class != provider.ErrTransient {
+		t.Errorf("class = %v, want ErrTransient so the proxy retries once the Secret lands", class)
+	}
+	if err := os.WriteFile(f.keyFile, keyPEM, 0o600); err != nil {
+		t.Fatal(err)
+	}
+	loaded, err := auth()
+	if err != nil {
+		t.Fatalf("loading after the key appeared: %v", err)
+	}
+	// Cached from here on: the file may vanish again (a Secret rotated out)
+	// without breaking connections that already work.
+	if err := os.Remove(f.keyFile); err != nil {
+		t.Fatal(err)
+	}
+	if again, err := auth(); err != nil || again != loaded {
+		t.Errorf("auth() = %v, %v; want the cached value reused", again, err)
+	}
+}
+
+func TestLazySSH_load_unparseableKeyIsPermanent(t *testing.T) {
+	t.Parallel()
+	f := newSSHFixture(t, true)
+	if err := os.WriteFile(f.keyFile, []byte("not a key"), 0o600); err != nil {
+		t.Fatal(err)
+	}
+	_, err := newLazySSH(f.cfg).load()
+	if classify(err) != provider.ErrPermanent {
+		t.Errorf("class = %v, want ErrPermanent: retrying cannot fix a key that does not parse", classify(err))
+	}
+}
+
+// The dialer must authenticate with the material the lazy loader produced,
+// not a stale or empty one. The echo socket is not a libvirt server, so the
+// call gets as far as the libvirt handshake and no further; what matters is
+// that it is past SSH by then.
+func TestLazySSH_dial_authenticatesOverSSH(t *testing.T) {
+	t.Parallel()
+	f := newSSHFixture(t, true)
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	_, err := newLazySSH(f.cfg).dial(ctx, f.host)
+	if err == nil {
+		t.Fatal("dial to an echo socket succeeded, want a libvirt handshake error")
+	}
+	if strings.Contains(err.Error(), "sshKeyFile") || strings.Contains(err.Error(), "ssh handshake") {
+		t.Errorf("dial failed before libvirt: %v", err)
+	}
+	if paths := f.server.requestedPaths(); len(paths) != 1 || paths[0] != f.host.Socket {
+		t.Errorf("server saw forward requests %v, want [%s]", paths, f.host.Socket)
+	}
+}
+
+// New must not fail on missing key material, or one misconfigured provider
+// aborts registry.Build and the manager exits before any controller starts.
+func TestNew_missingSSHMaterialStillBuilds(t *testing.T) {
+	t.Parallel()
+	p, err := New(context.Background(), provider.ProviderConfig{
+		Name: "kvm-home", Type: "kvm",
+		KVM: &provider.KVMConfig{
+			SSHKeyFile:     filepath.Join(t.TempDir(), "absent"),
+			KnownHostsFile: filepath.Join(t.TempDir(), "absent"),
+			Hosts:          []provider.KVMHost{{Name: "nas1", Address: "nas1:22", User: "virt", Bridge: "br0"}},
+		},
+	})
+	if err != nil || p == nil {
+		t.Fatalf("New() = %v, %v; want a usable provider", p, err)
+	}
+	if _, err := p.Get(context.Background(), "hosts/nas1/domains/proxy-x"); provider.Class(err) != provider.ErrTransient {
+		t.Errorf("Get class = %v, want ErrTransient", provider.Class(err))
+	}
+}
+
+// The health check reports where a connection broke, so every stage of a
+// real dial must come back tagged.
+func TestLazySSH_dial_tagsTheFailingStage(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name       string
+		mutate     func(t *testing.T, f *sshFixture)
+		want       dialStage
+		wantReason string
+	}{
+		{
+			name:       "key material missing",
+			mutate:     func(t *testing.T, f *sshFixture) { f.cfg.SSHKeyFile = filepath.Join(t.TempDir(), "absent") },
+			want:       stageSSHMaterial,
+			wantReason: reasonSSHMaterialUnavailable,
+		},
+		{
+			name: "nothing listening",
+			mutate: func(t *testing.T, f *sshFixture) {
+				ln, err := net.Listen("tcp", "127.0.0.1:0")
+				if err != nil {
+					t.Fatal(err)
+				}
+				f.host.Address = ln.Addr().String()
+				ln.Close()
+			},
+			want:       stageTCP,
+			wantReason: provider.ReasonUnreachable,
+		},
+		{
+			name: "host key not in known_hosts",
+			mutate: func(t *testing.T, f *sshFixture) {
+				impostor, _ := newKeyPair(t)
+				line := knownhosts.Line([]string{f.server.addr}, impostor.PublicKey())
+				if err := os.WriteFile(f.cfg.KnownHostsFile, []byte(line+"\n"), 0o600); err != nil {
+					t.Fatal(err)
+				}
+			},
+			want:       stageSSHHandshake,
+			wantReason: reasonSSHHandshakeFailed,
+		},
+		{
+			name: "client key rejected",
+			mutate: func(t *testing.T, f *sshFixture) {
+				_, otherPEM := newKeyPair(t)
+				if err := os.WriteFile(f.keyFile, otherPEM, 0o600); err != nil {
+					t.Fatal(err)
+				}
+			},
+			want:       stageSSHHandshake,
+			wantReason: reasonSSHHandshakeFailed,
+		},
+		{
+			name:       "libvirt socket missing",
+			mutate:     func(_ *testing.T, f *sshFixture) { f.host.Socket = "/nonexistent/libvirt-sock" },
+			want:       stageLibvirtConnect,
+			wantReason: reasonLibvirtConnectFailed,
+		},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			f := newSSHFixture(t, true)
+			tc.mutate(t, f)
+			ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+			defer cancel()
+
+			_, err := newLazySSH(f.cfg).dial(ctx, f.host)
+
+			var de *dialError
+			if !errors.As(err, &de) || de.stage != tc.want {
+				t.Fatalf("dial() error = %v, want a %s dialError", err, tc.want)
+			}
+			if got := checkReason(err); got != tc.wantReason {
+				t.Errorf("checkReason() = %s, want %s", got, tc.wantReason)
+			}
+		})
+	}
+}
+
 func TestKeepalive_closesClientWhenServerStalls(t *testing.T) {
 	t.Parallel()
 	f := newSSHFixture(t, false)
-	auth, err := newSSHAuth(f.cfg)
-	if err != nil {
-		t.Fatalf("newSSHAuth() error = %v", err)
-	}
-	client, err := auth.dialSSH(context.Background(), f.host)
+	client, err := mustSSHAuth(t, f.cfg).dialSSH(context.Background(), f.host)
 	if err != nil {
 		t.Fatalf("dialSSH() error = %v", err)
 	}
@@ -552,11 +809,7 @@ func TestKeepalive_closesClientWhenServerStalls(t *testing.T) {
 func TestKeepalive_keepsHealthyClientOpen(t *testing.T) {
 	t.Parallel()
 	f := newSSHFixture(t, true)
-	auth, err := newSSHAuth(f.cfg)
-	if err != nil {
-		t.Fatalf("newSSHAuth() error = %v", err)
-	}
-	client, err := auth.dialSSH(context.Background(), f.host)
+	client, err := mustSSHAuth(t, f.cfg).dialSSH(context.Background(), f.host)
 	if err != nil {
 		t.Fatalf("dialSSH() error = %v", err)
 	}
@@ -613,20 +866,7 @@ func TestHostPool_call_noTracingBelowV1(t *testing.T) {
 	}
 }
 
-func TestHostPool_call_logsDropReason(t *testing.T) {
-	t.Parallel()
-	t.Run("transport error", func(t *testing.T) {
-		t.Parallel()
-		d := &fakeDialer{}
-		p := newHostPool(d.dial)
-		log, snapshot := captureLogger(1)
-		_ = p.call(context.Background(), log, testHost, func(libvirtAPI) error { return io.ErrUnexpectedEOF })
-		joined := strings.Join(snapshot(), "\n")
-		if !strings.Contains(joined, `"msg"="libvirt connection dropped"`) || !strings.Contains(joined, `"reason"="transport error"`) {
-			t.Errorf("output missing the drop line:\n%s", joined)
-		}
-	})
-	t.Run("context cancelled", func(t *testing.T) {
+func TestHostPool_call_logsCancelDropReason(t *testing.T) {
 	t.Parallel()
 	d := &fakeDialer{}
 	p := newHostPool(d.dial)
@@ -650,5 +890,4 @@ func TestHostPool_call_logsDropReason(t *testing.T) {
 	if strings.Contains(joined, `"reason"="transport error"`) {
 		t.Errorf("cancel was also reported as a transport error:\n%s", joined)
 	}
-	})
 }

internal/provider/kvm/errors.go

@@ -19,10 +19,8 @@ import (
 // function to recognise. libvirt exposes no quota signal, so nothing maps
 // to QuotaExceeded.
 func classify(err error) error {
-	for _, class := range []error{provider.ErrNotFound, provider.ErrQuotaExceeded, provider.ErrPermanent, provider.ErrTransient} {
-		if errors.Is(err, class) {
-			return class
-		}
+	if isClassified(err) {
+		return provider.Class(err)
 	}
 	var lerr libvirt.Error
 	if !errors.As(err, &lerr) {
@@ -51,6 +49,18 @@ func isLibvirtError(err error) bool {
 	return errors.As(err, &lerr)
 }
 
+// isClassified reports whether err already carries one of the provider's
+// sentinels, i.e. the package reached a verdict itself rather than hitting
+// a transport failure.
+func isClassified(err error) bool {
+	for _, class := range []error{provider.ErrNotFound, provider.ErrQuotaExceeded, provider.ErrPermanent, provider.ErrTransient} {
+		if errors.Is(err, class) {
+			return true
+		}
+	}
+	return false
+}
+
 func hasCode(err error, code libvirt.ErrorNumber) bool {
 	var lerr libvirt.Error
 	return errors.As(err, &lerr) && libvirt.ErrorNumber(lerr.Code) == code

internal/provider/kvm/fake_test.go

@@ -28,10 +28,15 @@ type fakeAPI struct {
 	// uploadHook, when set, sees every StorageVolUpload before the fake
 	// reads the stream; it can return an error to simulate a broken upload.
 	uploadHook func(vol string, r io.Reader) error
+	// hostname, libVersion, poolCapacity and poolStates answer the health
+	// check's RPCs. A pool missing from poolStates is running.
+	hostname     string
+	libVersion   uint64
+	poolCapacity uint64
+	poolStates   map[string]libvirt.StoragePoolState
 }
 
 type fakeVol struct {
-	name     string
 	format   string
 	capacity int64
 	backing  string
@@ -54,6 +59,11 @@ func newFakeAPI(pools ...string) *fakeAPI {
 		domains:  map[string]*fakeDomain{},
 		fail:     map[string]error{},
 		failOnce: map[string]error{},
+
+		hostname:     "fake-host",
+		libVersion:   10002000,
+		poolCapacity: 100 << 30,
+		poolStates:   map[string]libvirt.StoragePoolState{},
 	}
 	for _, p := range pools {
 		f.pools[p] = map[string]*fakeVol{}
@@ -94,7 +104,7 @@ func (f *fakeAPI) vol(pool, name string) *fakeVol {
 func (f *fakeAPI) putVol(pool, name, format string, data []byte) {
 	f.mu.Lock()
 	defer f.mu.Unlock()
-	f.pools[pool][name] = &fakeVol{name: name, format: format, capacity: int64(len(data)), data: data}
+	f.pools[pool][name] = &fakeVol{format: format, capacity: int64(len(data)), data: data}
 }
 
 func (f *fakeAPI) volNames(pool string) []string {
@@ -348,7 +358,7 @@ func (f *fakeAPI) StorageVolCreateXML(pool libvirt.StoragePool, raw string, _ li
 	if _, exists := vols[v.Name]; exists {
 		return libvirt.StorageVol{}, lvError(libvirt.ErrStorageVolExist, "storage volume '%s' exists already", v.Name)
 	}
-	fv := &fakeVol{name: v.Name, format: v.Target.Format.Type, capacity: v.Capacity.Value}
+	fv := &fakeVol{format: v.Target.Format.Type, capacity: v.Capacity.Value}
 	if v.BackingStore != nil {
 		fv.backing = v.BackingStore.Path
 	}
@@ -410,3 +420,46 @@ func (f *fakeAPI) StorageVolDelete(vol libvirt.StorageVol, _ libvirt.StorageVolD
 	delete(f.pools[vol.Pool], vol.Name)
 	return nil
 }
+
+// --- host ---
+
+func (f *fakeAPI) ConnectGetLibVersion() (uint64, error) {
+	f.mu.Lock()
+	defer f.mu.Unlock()
+	if err := f.enter("ConnectGetLibVersion"); err != nil {
+		return 0, err
+	}
+	return f.libVersion, nil
+}
+
+func (f *fakeAPI) ConnectGetHostname() (string, error) {
+	f.mu.Lock()
+	defer f.mu.Unlock()
+	if err := f.enter("ConnectGetHostname"); err != nil {
+		return "", err
+	}
+	return f.hostname, nil
+}
+
+// StoragePoolGetInfo reports the pool's state and poolCapacity minus the
+// volumes' capacities as available space.
+func (f *fakeAPI) StoragePoolGetInfo(pool libvirt.StoragePool) (uint8, uint64, uint64, uint64, error) {
+	f.mu.Lock()
+	defer f.mu.Unlock()
+	if err := f.enter("StoragePoolGetInfo"); err != nil {
+		return 0, 0, 0, 0, err
+	}
+	vols, ok := f.pools[pool.Name]
+	if !ok {
+		return 0, 0, 0, 0, lvError(libvirt.ErrNoStoragePool, "no pool %s", pool.Name)
+	}
+	var used uint64
+	for _, v := range vols {
+		used += uint64(v.capacity)
+	}
+	state := libvirt.StoragePoolRunning
+	if s, ok := f.poolStates[pool.Name]; ok {
+		state = s
+	}
+	return uint8(state), f.poolCapacity, used, f.poolCapacity - used, nil
+}

internal/provider/kvm/image.go

@@ -8,7 +8,6 @@ import (
 	"fmt"
 	"io"
 	"net/http"
-	"net/url"
 	"strings"
 	"sync"
 	"time"
@@ -23,10 +22,7 @@ import (
 // provider's images map (Key set) or a URL given directly.
 type imageSource struct {
 	Key string
-	URL      string
-	SHA256   string
-	Format   string
-	Firmware string
+	provider.KVMImage
 }
 
 // resolveImage maps placement.image to a source. Anything that is neither a
@@ -34,11 +30,10 @@ type imageSource struct {
 // surface, not retry.
 func resolveImage(cfg provider.KVMConfig, image string) (imageSource, error) {
 	if img, ok := cfg.Images[image]; ok {
-		return imageSource{Key: image, URL: img.URL, SHA256: img.SHA256, Format: img.Format, Firmware: img.Firmware}, nil
+		return imageSource{Key: image, KVMImage: img}, nil
 	}
-	u, err := url.Parse(image)
-	if err == nil && (u.Scheme == "http" || u.Scheme == "https") && u.Host != "" {
-		return imageSource{URL: image, Format: formatQcow2, Firmware: firmwareUEFI}, nil
+	if provider.IsHTTPURL(image) {
+		return imageSource{KVMImage: imageDefaults(provider.KVMImage{URL: image})}, nil
 	}
 	return imageSource{}, fmt.Errorf("%w: placement.image %q is neither a key in kvm.images nor an http(s) URL",
 		provider.ErrPermanent, image)

internal/provider/kvm/image_test.go

@@ -37,12 +37,11 @@ type imageServer struct {
 	status   int
 	noLength bool
 	delay    time.Duration
-	body     []byte
 }
 
 func newImageServer(t *testing.T) *imageServer {
 	t.Helper()
-	s := &imageServer{status: http.StatusOK, body: testImage}
+	s := &imageServer{status: http.StatusOK}
 	s.Server = httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
 		s.hits.Add(1)
 		time.Sleep(s.delay)
@@ -55,16 +54,16 @@ func newImageServer(t *testing.T) *imageServer {
 		if s.noLength {
 			w.(http.Flusher).Flush()
 		} else {
-			w.Header().Set("Content-Length", strconv.Itoa(len(s.body)))
+			w.Header().Set("Content-Length", strconv.Itoa(len(testImage)))
 		}
-		_, _ = w.Write(s.body)
+		_, _ = w.Write(testImage)
 	}))
 	t.Cleanup(s.Close)
 	return s
 }
 
 func testSource(srv *imageServer, sha string) imageSource {
-	return imageSource{Key: "debian-13", URL: srv.URL + "/debian.qcow2", SHA256: sha, Format: formatQcow2}
+	return imageSource{Key: "debian-13", KVMImage: provider.KVMImage{URL: srv.URL + "/debian.qcow2", SHA256: sha, Format: formatQcow2}}
 }
 
 func ensureTest(t *testing.T, api *fakeAPI, src imageSource) (string, error) {
@@ -88,9 +87,9 @@ func TestResolveImage(t *testing.T) {
 		want    imageSource
 		wantErr bool
 	}{
-		{"configured key", "debian-13", imageSource{Key: "debian-13", URL: "https://cloud.debian.org/d.qcow2", SHA256: "abc", Format: "qcow2", Firmware: "uefi"}, false},
-		{"direct https url", "https://x.example/img.qcow2", imageSource{URL: "https://x.example/img.qcow2", Format: "qcow2", Firmware: "uefi"}, false},
-		{"direct http url", "http://mirror/img.qcow2", imageSource{URL: "http://mirror/img.qcow2", Format: "qcow2", Firmware: "uefi"}, false},
+		{"configured key", "debian-13", imageSource{Key: "debian-13", KVMImage: provider.KVMImage{URL: "https://cloud.debian.org/d.qcow2", SHA256: "abc", Format: "qcow2", Firmware: "uefi"}}, false},
+		{"direct https url", "https://x.example/img.qcow2", imageSource{KVMImage: provider.KVMImage{URL: "https://x.example/img.qcow2", Format: "qcow2", Firmware: "uefi"}}, false},
+		{"direct http url", "http://mirror/img.qcow2", imageSource{KVMImage: provider.KVMImage{URL: "http://mirror/img.qcow2", Format: "qcow2", Firmware: "uefi"}}, false},
 		{"unknown key", "ubuntu", imageSource{}, true},
 		{"file url", "file:///tmp/x.qcow2", imageSource{}, true},
 		{"empty", "", imageSource{}, true},
@@ -100,9 +99,7 @@ func TestResolveImage(t *testing.T) {
 			t.Parallel()
 			got, err := resolveImage(cfg, tc.image)
 			if tc.wantErr {
-				if !errors.Is(err, provider.ErrPermanent) {
-					t.Fatalf("resolveImage(%q) error = %v, want ErrPermanent", tc.image, err)
-				}
+				wantClass(t, err, provider.ErrPermanent)
 				return
 			}
 			if err != nil {
@@ -117,16 +114,16 @@ func TestResolveImage(t *testing.T) {
 
 func TestImageSource_volumeNames(t *testing.T) {
 	t.Parallel()
-	keyed := imageSource{Key: "debian-13", URL: "https://a/b.qcow2", Format: "qcow2"}
+	keyed := imageSource{Key: "debian-13", KVMImage: provider.KVMImage{URL: "https://a/b.qcow2", Format: "qcow2"}}
 	if keyed.volumeName() != "base-debian-13.qcow2" || keyed.markerName() != "base-debian-13.qcow2.sha256" {
 		t.Errorf("keyed names = %s / %s", keyed.volumeName(), keyed.markerName())
 	}
-	direct := imageSource{URL: "https://a/b.qcow2", Format: "qcow2"}
+	direct := imageSource{KVMImage: provider.KVMImage{URL: "https://a/b.qcow2", Format: "qcow2"}}
 	name := direct.volumeName()
 	if !strings.HasPrefix(name, "base-") || !strings.HasSuffix(name, ".qcow2") || len(name) != len("base-")+16+len(".qcow2") {
 		t.Errorf("direct-url name = %q, want base-<16 hex>.qcow2", name)
 	}
-	other := imageSource{URL: "https://a/c.qcow2", Format: "qcow2"}
+	other := imageSource{KVMImage: provider.KVMImage{URL: "https://a/c.qcow2", Format: "qcow2"}}
 	if other.volumeName() == name {
 		t.Error("different URLs produced the same volume name")
 	}
@@ -181,12 +178,27 @@ func TestImageCache_ensure_noChecksumConfigured(t *testing.T) {
 	}
 }
 
-func TestImageCache_ensure_alreadyCached(t *testing.T) {
+func TestImageCache_ensure_existingVolumes(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name     string
+		base     []byte
+		marker   []byte
+		wantHits int32
+	}{
+		{"already cached", testImage, []byte(testImageSHA()), 0},
+		{"marker missing is a partial upload", []byte("half an image"), nil, 1},
+		{"configured checksum changed", []byte("an older release"), []byte(strings.Repeat("0", 64)), 1},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
 			t.Parallel()
 			srv := newImageServer(t)
 			api := newFakeAPI("default")
-	api.putVol("default", "base-debian-13.qcow2", "qcow2", testImage)
-	api.putVol("default", "base-debian-13.qcow2.sha256", "raw", []byte(testImageSHA()))
+			api.putVol("default", "base-debian-13.qcow2", "qcow2", tc.base)
+			if tc.marker != nil {
+				api.putVol("default", "base-debian-13.qcow2.sha256", "raw", tc.marker)
+			}
 			path, err := ensureTest(t, api, testSource(srv, testImageSHA()))
 			if err != nil {
 				t.Fatalf("ensure() error = %v", err)
@@ -194,44 +206,19 @@ func TestImageCache_ensure_alreadyCached(t *testing.T) {
 			if path != "/var/lib/libvirt/images/base-debian-13.qcow2" {
 				t.Errorf("path = %q", path)
 			}
-	if srv.hits.Load() != 0 {
-		t.Errorf("http hits = %d, want 0 for a cached image", srv.hits.Load())
+			if srv.hits.Load() != tc.wantHits {
+				t.Errorf("http hits = %d, want %d", srv.hits.Load(), tc.wantHits)
 			}
-	if api.hasCall("StorageVolCreateXML") || api.hasCall("StorageVolDelete") {
-		t.Errorf("cached image was touched: calls = %v", api.calls)
-	}
-}
-
-func TestImageCache_ensure_markerMissingRedownloads(t *testing.T) {
-	t.Parallel()
-	srv := newImageServer(t)
-	api := newFakeAPI("default")
-	api.putVol("default", "base-debian-13.qcow2", "qcow2", []byte("half an image"))
-	if _, err := ensureTest(t, api, testSource(srv, testImageSHA())); err != nil {
-		t.Fatalf("ensure() error = %v", err)
-	}
-	if srv.hits.Load() != 1 {
-		t.Errorf("http hits = %d, want 1: a base without its marker is a partial upload", srv.hits.Load())
+			if touched := api.hasCall("StorageVolCreateXML") || api.hasCall("StorageVolDelete"); touched != (tc.wantHits > 0) {
+				t.Errorf("volumes touched = %v, want %v: calls = %v", touched, tc.wantHits > 0, api.calls)
 			}
 			if base := api.vol("default", "base-debian-13.qcow2"); base == nil || !bytes.Equal(base.data, testImage) {
-		t.Error("partial base volume was not replaced")
-	}
-}
-
-func TestImageCache_ensure_checksumChangedRedownloads(t *testing.T) {
-	t.Parallel()
-	srv := newImageServer(t)
-	api := newFakeAPI("default")
-	api.putVol("default", "base-debian-13.qcow2", "qcow2", []byte("an older release"))
-	api.putVol("default", "base-debian-13.qcow2.sha256", "raw", []byte(strings.Repeat("0", 64)))
-	if _, err := ensureTest(t, api, testSource(srv, testImageSHA())); err != nil {
-		t.Fatalf("ensure() error = %v", err)
-	}
-	if srv.hits.Load() != 1 {
-		t.Errorf("http hits = %d, want 1 after the configured checksum changed", srv.hits.Load())
+				t.Error("base volume does not hold the verified image")
 			}
 			if marker := api.vol("default", "base-debian-13.qcow2.sha256"); marker == nil || string(marker.data) != testImageSHA() {
-		t.Errorf("marker not updated: %+v", marker)
+				t.Errorf("marker = %+v, want the verified digest", marker)
+			}
+		})
 	}
 }
 
@@ -240,9 +227,7 @@ func TestImageCache_ensure_checksumMismatchIsPermanent(t *testing.T) {
 	srv := newImageServer(t)
 	api := newFakeAPI("default")
 	_, err := ensureTest(t, api, testSource(srv, strings.Repeat("a", 64)))
-	if !errors.Is(err, provider.ErrPermanent) {
-		t.Fatalf("ensure() error = %v, want ErrPermanent", err)
-	}
+	wantClass(t, err, provider.ErrPermanent)
 	if names := api.volNames("default"); len(names) != 0 {
 		t.Errorf("volumes left behind after a checksum mismatch: %v", names)
 	}
@@ -283,8 +268,9 @@ func TestImageCache_ensure_noContentLengthIsPermanent(t *testing.T) {
 	srv.noLength = true
 	api := newFakeAPI("default")
 	_, err := ensureTest(t, api, testSource(srv, ""))
-	if !errors.Is(err, provider.ErrPermanent) || !strings.Contains(err.Error(), "Content-Length") {
-		t.Fatalf("ensure() error = %v, want ErrPermanent about Content-Length", err)
+	wantClass(t, err, provider.ErrPermanent)
+	if !strings.Contains(err.Error(), "Content-Length") {
+		t.Errorf("ensure() error = %v, want one about Content-Length", err)
 	}
 }
 

internal/provider/kvm/kvm.go

@@ -26,10 +26,10 @@ const (
 	defaultDiskSizeGB = 10
 	defaultSocket     = "/var/run/libvirt/libvirt-sock"
 	defaultPool       = "default"
-	formatQcow2       = "qcow2"
-	formatRaw         = "raw"
-	firmwareUEFI      = "uefi"
-	firmwareBIOS      = "bios"
+	formatQcow2       = provider.KVMFormatQcow2
+	formatRaw         = provider.KVMFormatRaw
+	firmwareUEFI      = provider.KVMFirmwareUEFI
+	firmwareBIOS      = provider.KVMFirmwareBIOS
 
 	// undefineFlags removes everything libvirt may keep alongside a domain
 	// definition, so an undefine never fails on leftover state.
@@ -42,18 +42,9 @@ func withDefaults(cfg provider.KVMConfig) provider.KVMConfig {
 	if cfg.DiskSizeGB == 0 {
 		cfg.DiskSizeGB = defaultDiskSizeGB
 	}
-	if cfg.VendorData == "" {
-		cfg.VendorData = defaultVendorData
-	}
 	images := make(map[string]provider.KVMImage, len(cfg.Images))
 	for k, img := range cfg.Images {
-		if img.Format == "" {
-			img.Format = formatQcow2
-		}
-		if img.Firmware == "" {
-			img.Firmware = firmwareUEFI
-		}
-		images[k] = img
+		images[k] = imageDefaults(img)
 	}
 	cfg.Images = images
 	hosts := make([]provider.KVMHost, len(cfg.Hosts))
@@ -79,30 +70,41 @@ type Provider struct {
 	now    func() time.Time
 }
 
+func imageDefaults(img provider.KVMImage) provider.KVMImage {
+	if img.Format == "" {
+		img.Format = formatQcow2
+	}
+	if img.Firmware == "" {
+		img.Firmware = firmwareUEFI
+	}
+	return img
+}
+
 // New builds a Provider that reaches its hosts over SSH. The key and
-// known_hosts files are read here so a bad path fails at startup rather
-// than on the first Create. Satisfies registry.Constructor.
-func New(_ context.Context, pc provider.ProviderConfig) (provider.Provider, error) {
+// known_hosts files are tried once here and kept when usable, but a failure
+// is only a warning, retried on the first connection: they come from an
+// optional Secret mount, and a constructor that failed without it would
+// abort the whole provider set and crash-loop the manager, taking the
+// kubernetes and gcp providers down with it. Satisfies registry.Constructor.
+func New(ctx context.Context, pc provider.ProviderConfig) (provider.Provider, error) {
 	cfg := provider.KVMConfig{}
 	if pc.KVM != nil {
 		cfg = *pc.KVM
 	}
-	auth, err := newSSHAuth(cfg)
-	if err != nil {
-		return nil, fmt.Errorf("kvm provider %q: %w", pc.Name, err)
+	loader := newLazySSH(cfg)
+	if _, err := loader.load(); err != nil {
+		logf.FromContext(ctx).WithName("kvm").WithValues("provider", pc.Name).Info(
+			"WARNING: kvm ssh material is not usable; proxies on this provider will fail until it is",
+			"error", err.Error())
 	}
-	return newWithDialer(pc, auth.dialer(), nil), nil
+	return newWithDialer(pc.Name, cfg, loader.dial, nil), nil
 }
 
 // newWithDialer is the test seam: the dialer decides what libvirtAPI each
 // host resolves to, and the HTTP client is what image downloads go through.
-func newWithDialer(pc provider.ProviderConfig, dial hostDialer, httpClient *http.Client) *Provider {
-	cfg := provider.KVMConfig{}
-	if pc.KVM != nil {
-		cfg = *pc.KVM
-	}
+func newWithDialer(name string, cfg provider.KVMConfig, dial hostDialer, httpClient *http.Client) *Provider {
 	return &Provider{
-		name:   pc.Name,
+		name:   name,
 		cfg:    withDefaults(cfg),
 		hosts:  newHostPool(dial),
 		images: newImageCache(httpClient),
@@ -249,7 +251,11 @@ func (p *Provider) writeSeed(log logr.Logger, api libvirtAPI, pool libvirt.Stora
 	if err := deleteIfExists(api, pool, seedName); err != nil {
 		return libvirt.StorageVol{}, err
 	}
-	iso, err := buildSeedISO(name, userData, p.cfg.VendorData)
+	vendorData, err := buildVendorData(p.cfg.VendorData, p.cfg.InstallGuestAgent == nil || *p.cfg.InstallGuestAgent)
+	if err != nil {
+		return libvirt.StorageVol{}, fmt.Errorf("%w: %v", provider.ErrPermanent, err)
+	}
+	iso, err := buildSeedISO(name, userData, vendorData)
 	if err != nil {
 		return libvirt.StorageVol{}, fmt.Errorf("%w: %v", provider.ErrPermanent, err)
 	}
@@ -450,7 +456,8 @@ func mapState(s libvirt.DomainState, ip string) provider.InstanceState {
 // domain is not destroyed again, a missing domain is not undefined, and
 // the per-VM volumes are removed by name even when the domain is already
 // gone, so a crash between volume creation and define is cleaned up on the
-// next reconcile.
+// next reconcile. A providerID naming a host that is no longer configured
+// is a no-op success rather than an error.
 func (p *Provider) Delete(ctx context.Context, providerID string) error {
 	hostName, name, err := parseProviderID(providerID)
 	if err != nil {
@@ -458,7 +465,14 @@ func (p *Provider) Delete(ctx context.Context, providerID string) error {
 	}
 	host, err := p.host(hostName)
 	if err != nil {
-		return p.wrapErr("delete", providerID, err)
+		// The host is no longer configured, so this operator has no way to
+		// reach the domain and nothing left to clean. Same reasoning as
+		// removeVolumes over a missing pool: refusing would pin the Proxy's
+		// finalizer. Whatever is still running there is the admin's to remove,
+		// and orphan GC cannot see it either — see docs/kvm-provider.md.
+		p.logger(ctx).Info("host is not configured; delete has nothing to clean up",
+			"host", hostName, "providerID", providerID)
+		return nil
 	}
 	log := p.logger(ctx).WithValues("host", host.Name, "name", name)
 	err = p.hosts.call(ctx, log, host, func(api libvirtAPI) error {

internal/provider/kvm/kvm_integration_test.go

@@ -347,6 +347,36 @@ func TestKVM_lifecycle(t *testing.T) {
 	env.keep = false
 }
 
+// TestKVMCheck_realHost runs the provider health check against the host.
+// It creates nothing: it must reach the host and find the pool running.
+func TestKVMCheck_realHost(t *testing.T) {
+	env := loadIntegrationEnv(t)
+	log := integrationLogger(t, env.logLevel)
+	ctx, cancel := context.WithTimeout(logr.NewContext(context.Background(), log), time.Minute)
+	defer cancel()
+
+	prov, err := New(ctx, env.cfg)
+	if err != nil {
+		t.Fatalf("New: %v", err)
+	}
+	got := prov.(*Provider).Check(ctx)
+
+	if len(got) != 1 {
+		t.Fatalf("Check returned %d targets, want 1", len(got))
+	}
+	ts := got[0]
+	t.Logf("%s (%s): ready=%v reason=%s latency=%s details=%v",
+		ts.Name, ts.Address, ts.Ready, ts.Reason, ts.Latency.Round(time.Millisecond), ts.Details)
+	if !ts.Ready {
+		t.Fatalf("check failed: %s: %s", ts.Reason, ts.Message)
+	}
+	for _, key := range []string{"hostname", "libvirtVersion", "poolState", "poolAvailable"} {
+		if ts.Details[key] == "" {
+			t.Errorf("details[%s] is empty, want what the host reported", key)
+		}
+	}
+}
+
 // waitForRunning polls Get until the instance is Running with an IP or ctx
 // ends. Transient errors are retried; anything else fails the test.
 func waitForRunning(ctx context.Context, t *testing.T, p *Provider, id string) *provider.Instance {

internal/provider/kvm/kvm_test.go

@@ -30,8 +30,8 @@ func TestWithDefaults(t *testing.T) {
 	if cfg.DiskSizeGB != defaultDiskSizeGB {
 		t.Errorf("diskSizeGb = %d, want %d", cfg.DiskSizeGB, defaultDiskSizeGB)
 	}
-	if cfg.VendorData != defaultVendorData {
-		t.Errorf("vendorData = %q, want the default", cfg.VendorData)
+	if cfg.VendorData != "" {
+		t.Errorf("vendorData = %q, want none: the guest-agent bootstrap is merged into the user-data, not defaulted here", cfg.VendorData)
 	}
 	if cfg.Images["debian"].Format != "qcow2" || cfg.Images["raw"].Format != "raw" {
 		t.Errorf("image formats = %q/%q, want qcow2 default and raw kept", cfg.Images["debian"].Format, cfg.Images["raw"].Format)
@@ -95,7 +95,8 @@ func testConfig(hosts ...string) provider.ProviderConfig {
 }
 
 func newTestProvider(fakes hostFakes, hosts ...string) *Provider {
-	p := newWithDialer(testConfig(hosts...), fakes.dial, nil)
+	pc := testConfig(hosts...)
+	p := newWithDialer(pc.Name, *pc.KVM, fakes.dial, nil)
 	p.now = func() time.Time { return testNow }
 	return p
 }
@@ -103,7 +104,7 @@ func newTestProvider(fakes hostFakes, hosts ...string) *Provider {
 // cacheImage pre-populates the pool with the configured image and its
 // marker so Create never touches the network.
 func cacheImage(api *fakeAPI) {
-	src := imageSource{Key: "debian-13", URL: testImageURL, Format: formatQcow2}
+	src := imageSource{Key: "debian-13", KVMImage: provider.KVMImage{URL: testImageURL, Format: formatQcow2}}
 	api.putVol("default", src.volumeName(), formatQcow2, []byte("qcow2"))
 	api.putVol("default", src.markerName(), formatRaw, []byte(testImageSHA()))
 }
@@ -187,6 +188,13 @@ func assertCreated(t *testing.T, api *fakeAPI) {
 	if seed.format != formatRaw || len(seed.data) == 0 || int64(len(seed.data)) != seed.capacity {
 		t.Errorf("seed = format %s, %d bytes, capacity %d; want raw with capacity == uploaded size", seed.format, len(seed.data), seed.capacity)
 	}
+	_, files := readSeed(t, seed.data)
+	if files["user-data"] != testCloudInit {
+		t.Errorf("seed user-data = %q, want the Proxy's cloud-init verbatim", files["user-data"])
+	}
+	if !strings.Contains(files["vendor-data"], "qemu-guest-agent") {
+		t.Errorf("seed vendor-data = %q, want the guest-agent script", files["vendor-data"])
+	}
 }
 
 func TestCreate_freshHost(t *testing.T) {
@@ -234,7 +242,7 @@ func TestCreate_fetchesImageOnFirstUse(t *testing.T) {
 	api := newFakeAPI("default")
 	cfg := testConfig("nas1")
 	cfg.KVM.Images["debian-13"] = provider.KVMImage{URL: srv.URL + "/debian.qcow2", SHA256: testImageSHA()}
-	p := newWithDialer(cfg, hostFakes{"nas1": api}.dial, nil)
+	p := newWithDialer(cfg.Name, *cfg.KVM, hostFakes{"nas1": api}.dial, nil)
 
 	if _, err := p.Create(context.Background(), testRequest()); err != nil {
 		t.Fatalf("Create: %v", err)
@@ -472,28 +480,25 @@ func TestCreate_hostFailures(t *testing.T) {
 	t.Parallel()
 	cases := []struct {
 		name  string
-		setup func() (hostFakes, string)
+		setup func() hostFakes
 		class error
 	}{
 		{
 			name:  "host unreachable",
-			setup: func() (hostFakes, string) { return hostFakes{}, "nas1" },
+			setup: func() hostFakes { return hostFakes{} },
 			class: provider.ErrTransient,
 		},
 		{
 			name:  "pool missing",
-			setup: func() (hostFakes, string) {
-				api := newFakeAPI("other")
-				return hostFakes{"nas1": api}, "nas1"
-			},
+			setup: func() hostFakes { return hostFakes{"nas1": newFakeAPI("other")} },
 			class: provider.ErrPermanent,
 		},
 		{
 			name: "domain xml rejected",
-			setup: func() (hostFakes, string) {
+			setup: func() hostFakes {
 				api := readyHost()
 				api.fail["DomainDefineXML"] = lvError(libvirt.ErrConfigUnsupported, "no q35 here")
-				return hostFakes{"nas1": api}, "nas1"
+				return hostFakes{"nas1": api}
 			},
 			class: provider.ErrPermanent,
 		},
@@ -501,8 +506,7 @@ func TestCreate_hostFailures(t *testing.T) {
 	for _, tc := range cases {
 		t.Run(tc.name, func(t *testing.T) {
 			t.Parallel()
-			fakes, host := tc.setup()
-			p := newTestProvider(fakes, host)
+			p := newTestProvider(tc.setup(), "nas1")
 			_, err := p.Create(context.Background(), testRequest())
 			wantClass(t, err, tc.class)
 		})
@@ -722,6 +726,28 @@ func TestDelete_missingPoolIsNotAnError(t *testing.T) {
 	}
 }
 
+// A host dropped from the config leaves the operator no way to reach the
+// domain, so Delete has nothing left to clean. Failing instead would pin
+// the Proxy's finalizer forever, since the reconciler's delete path cannot
+// make progress without a Delete that can succeed.
+func TestDelete_unconfiguredHostIsNotAnError(t *testing.T) {
+	t.Parallel()
+	api := readyHost()
+	managedDomain(t, api, testName, testUID, true, nil)
+	p := newTestProvider(hostFakes{"nas1": api}, "nas1")
+
+	if err := p.Delete(context.Background(), "hosts/nas9/domains/"+testName); err != nil {
+		t.Fatalf("Delete over an unconfigured host: %v", err)
+	}
+	if api.domain(testName) == nil {
+		t.Error("domain on the configured host was touched; only the named host's state may change")
+	}
+	// Get still refuses, so the live path cannot mistake a config typo for a
+	// missing VM and recreate it somewhere else.
+	_, err := p.Get(context.Background(), "hosts/nas9/domains/"+testName)
+	wantClass(t, err, provider.ErrPermanent)
+}
+
 func TestDelete_errors(t *testing.T) {
 	t.Parallel()
 	cases := []struct {
@@ -731,7 +757,6 @@ func TestDelete_errors(t *testing.T) {
 		class error
 	}{
 		{"malformed id", "nope", func(*testing.T) hostFakes { return hostFakes{"nas1": readyHost()} }, provider.ErrPermanent},
-		{"host not configured", "hosts/nas9/domains/x", func(*testing.T) hostFakes { return hostFakes{"nas1": readyHost()} }, provider.ErrPermanent},
 		{"host unreachable", "hosts/nas1/domains/x", func(*testing.T) hostFakes { return hostFakes{} }, provider.ErrTransient},
 		{
 			"undefine fails", "hosts/nas1/domains/" + testName,
@@ -880,15 +905,6 @@ func TestMapState(t *testing.T) {
 	}
 }
 
-func TestNew_badKeyPathFailsFast(t *testing.T) {
-	t.Parallel()
-	cfg := testConfig("nas1")
-	cfg.KVM.SSHKeyFile = t.TempDir() + "/missing"
-	if _, err := New(context.Background(), cfg); err == nil {
-		t.Fatal("New succeeded with a missing key file")
-	}
-}
-
 // runAllOps drives one full lifecycle over a ready fake host so logging
 // tests see every provider method and every libvirt call.
 func runAllOps(t *testing.T, ctx context.Context, req provider.CreateRequest) {
@@ -956,7 +972,7 @@ func TestLogging_verbosityTiers(t *testing.T) {
 			ctx, snapshot := captureContext(tc.verbosity)
 			req := testRequest()
 			const sentinel = "SENTINEL-cloud-init-must-never-be-logged"
-			req.CloudInit = "#cloud-config\n" + sentinel + "\n"
+			req.CloudInit = "#cloud-config\nhostname: " + sentinel + "\n"
 
 			runAllOps(t, ctx, req)
 

internal/provider/kvm/metadata.go

@@ -38,14 +38,12 @@ func newInstanceMetadata(uid string, createdAt time.Time) instanceMetadata {
 func parseInstanceMetadata(s string) (instanceMetadata, error) {
 	var raw struct {
 		XMLName xml.Name `xml:"instance"`
-		Managed   string   `xml:"managed"`
-		UID       string   `xml:"uid"`
-		CreatedAt string   `xml:"createdAt"`
+		instanceMetadata
 	}
 	if err := xml.Unmarshal([]byte(s), &raw); err != nil {
 		return instanceMetadata{}, fmt.Errorf("parsing domain metadata: %w", err)
 	}
-	return instanceMetadata{Managed: raw.Managed, UID: raw.UID, CreatedAt: raw.CreatedAt}, nil
+	return raw.instanceMetadata, nil
 }
 
 func (m instanceMetadata) managed() bool {

internal/provider/kvm/metadata_test.go

@@ -29,47 +29,37 @@ func TestInstanceMetadata_roundTrip(t *testing.T) {
 	}
 }
 
-func TestParseInstanceMetadata_libvirtPrefixedForm(t *testing.T) {
-	t.Parallel()
-	// libvirt re-serialises the stored element with whatever prefix it
-	// chooses; only the namespace URI is contractual.
-	in := `<op:instance xmlns:op="` + metadataURI + `">` +
-		`<op:managed>true</op:managed><op:uid>uid-2</op:uid>` +
-		`<op:createdAt>2026-09-04T10:00:00Z</op:createdAt></op:instance>`
-	m, err := parseInstanceMetadata(in)
-	if err != nil {
-		t.Fatalf("parseInstanceMetadata() error = %v", err)
-	}
-	if !m.managed() || m.UID != "uid-2" || m.createdAt().Hour() != 10 {
-		t.Errorf("parsed = %+v", m)
-	}
-}
-
-func TestParseInstanceMetadata_libvirtBareForm(t *testing.T) {
-	t.Parallel()
-	// What libvirt 11.6 actually returns from DomainGetMetadata: the element
-	// selected by our URI, serialised without any namespace declaration.
-	in := `<instance><managed>true</managed><uid>uid-3</uid>` +
-		`<createdAt>2026-09-10T07:30:00Z</createdAt></instance>`
-	m, err := parseInstanceMetadata(in)
-	if err != nil {
-		t.Fatalf("parseInstanceMetadata() error = %v", err)
-	}
-	if !m.managed() || m.UID != "uid-3" || m.createdAt().Minute() != 30 {
-		t.Errorf("parsed = %+v", m)
-	}
-}
-
-func TestParseInstanceMetadata_notManaged(t *testing.T) {
+func TestParseInstanceMetadata(t *testing.T) {
 	t.Parallel()
 	tests := []struct {
 		name        string
 		in          string
 		wantErr     bool
+		wantManaged bool
+		wantUID     string
+		wantCreated time.Time
 	}{
-		{"managed false", `<instance xmlns="` + metadataURI + `"><managed>false</managed><uid>u</uid></instance>`, false},
-		{"foreign element", `<foo xmlns="urn:other"><managed>true</managed></foo>`, true},
-		{"garbage", `not xml`, true},
+		{
+			// libvirt re-serialises the stored element with whatever prefix it
+			// chooses; only the namespace URI is contractual.
+			name: "libvirt prefixed form",
+			in: `<op:instance xmlns:op="` + metadataURI + `">` +
+				`<op:managed>true</op:managed><op:uid>uid-2</op:uid>` +
+				`<op:createdAt>2026-09-04T10:00:00Z</op:createdAt></op:instance>`,
+			wantManaged: true, wantUID: "uid-2", wantCreated: time.Date(2026, 9, 4, 10, 0, 0, 0, time.UTC),
+		},
+		{
+			// What libvirt 11.6 actually returns from DomainGetMetadata: the
+			// element selected by our URI, serialised without any namespace
+			// declaration.
+			name: "libvirt bare form",
+			in: `<instance><managed>true</managed><uid>uid-3</uid>` +
+				`<createdAt>2026-09-10T07:30:00Z</createdAt></instance>`,
+			wantManaged: true, wantUID: "uid-3", wantCreated: time.Date(2026, 9, 10, 7, 30, 0, 0, time.UTC),
+		},
+		{name: "managed false", in: `<instance xmlns="` + metadataURI + `"><managed>false</managed><uid>u</uid></instance>`, wantUID: "u"},
+		{name: "foreign element", in: `<foo xmlns="urn:other"><managed>true</managed></foo>`, wantErr: true},
+		{name: "garbage", in: `not xml`, wantErr: true},
 	}
 	for _, tc := range tests {
 		t.Run(tc.name, func(t *testing.T) {
@@ -78,8 +68,8 @@ func TestParseInstanceMetadata_notManaged(t *testing.T) {
 			if (err != nil) != tc.wantErr {
 				t.Fatalf("error = %v, wantErr %v", err, tc.wantErr)
 			}
-			if m.managed() {
-				t.Errorf("managed() = true for %q", tc.in)
+			if m.managed() != tc.wantManaged || m.UID != tc.wantUID || !m.createdAt().Equal(tc.wantCreated) {
+				t.Errorf("parsed = %+v, want managed=%v uid=%q createdAt=%v", m, tc.wantManaged, tc.wantUID, tc.wantCreated)
 			}
 		})
 	}

internal/provider/kvm/seed.go

@@ -3,6 +3,9 @@ package kvm
 import (
 	"bytes"
 	"fmt"
+	"io"
+	"mime/multipart"
+	"net/textproto"
 	"strings"
 
 	"github.com/kdomanski/iso9660"
@@ -11,20 +14,97 @@ import (
 // seedLabel is the volume label cloud-init's NoCloud datasource looks for.
 const seedLabel = "cidata"
 
-// defaultVendorData is the cloud-init vendor-data shipped on every seed
-// unless the config overrides it. cloud-init merges it under the Proxy's own
-// user-data, so stock images grow a guest agent (which lets the provider
-// read the VM's IP before ARP sees traffic) without the user's cloud-init
-// having to know about it.
-const defaultVendorData = `#cloud-config
-packages:
-  - qemu-guest-agent
-runcmd:
-  - [systemctl, enable, --now, qemu-guest-agent]
+// guestAgentScript installs and starts the QEMU guest agent, which is how
+// the provider learns a VM's address before the host's ARP table has seen
+// any traffic from it.
+//
+// It ships as vendor-data, and as a *script* rather than a cloud-config,
+// because that is the only shape cloud-init cannot drop. Vendor-data
+// cloud-config is merged under the Proxy's own user-data with mergers that
+// keep the user-data's value for every key both documents define and never
+// append lists, so a Proxy carrying its own packages: silently lost the
+// agent. The strategy cannot be overridden from either document: merge_how
+// in user-data governs only the first source, and merge_how in vendor-data
+// is consumed and discarded before the document is stored. A vendor-data
+// script bypasses all of it — cloud-init gives vendor-data its own handler
+// chain, writes scripts to a separate directory and runs them from
+// cc_scripts_vendor, after the user's packages are installed and before the
+// user's runcmd.
+const guestAgentScript = `#!/bin/sh
+# Added by egress-proxies-operator so the hypervisor can report this VM's
+# address. Set installGuestAgent: false on the provider to leave it out.
+set -e
+if ! command -v qemu-ga >/dev/null 2>&1 && [ ! -x /usr/sbin/qemu-ga ]; then
+	if command -v apt-get >/dev/null 2>&1; then
+		apt-get update
+		DEBIAN_FRONTEND=noninteractive apt-get install -y qemu-guest-agent
+	elif command -v dnf >/dev/null 2>&1; then
+		dnf install -y qemu-guest-agent
+	elif command -v yum >/dev/null 2>&1; then
+		yum install -y qemu-guest-agent
+	elif command -v zypper >/dev/null 2>&1; then
+		zypper --non-interactive install qemu-guest-agent
+	elif command -v apk >/dev/null 2>&1; then
+		apk add qemu-guest-agent
+	else
+		echo "no known package manager; install qemu-guest-agent in the image" >&2
+		exit 1
+	fi
+fi
+if command -v systemctl >/dev/null 2>&1; then
+	systemctl enable --now qemu-guest-agent
+elif command -v rc-update >/dev/null 2>&1; then
+	rc-update add qemu-guest-agent default
+	rc-service qemu-guest-agent start
+fi
 `
 
-// buildSeedISO renders a NoCloud seed: user-data verbatim, vendor-data as
-// given, and a meta-data whose instance-id is the deterministic VM name, so
+// buildVendorData renders the seed's vendor-data. With nothing configured
+// it is the guest-agent script alone; with both, a MIME multipart carrying
+// the configured document and the script as separate parts, since
+// vendor-data is a single file and the two must not be concatenated.
+func buildVendorData(configured string, installAgent bool) (string, error) {
+	switch {
+	case !installAgent:
+		return configured, nil
+	case strings.TrimSpace(configured) == "":
+		return guestAgentScript, nil
+	default:
+		return multipartVendorData(configured, guestAgentScript)
+	}
+}
+
+// multipartVendorData packs several documents into one vendor-data blob.
+// cloud-init decides a payload is multipart by finding "mime-version:"
+// within its first 4096 bytes, so those headers lead. The configured
+// document goes in as text/plain on purpose: cloud-init re-sniffs that type
+// from the content's own header line, which keeps this code out of the
+// business of classifying what the user wrote.
+func multipartVendorData(configured, script string) (string, error) {
+	var body bytes.Buffer
+	w := multipart.NewWriter(&body)
+	parts := []struct{ contentType, content string }{
+		{"text/plain", configured},
+		{"text/x-shellscript", script},
+	}
+	for _, p := range parts {
+		part, err := w.CreatePart(textproto.MIMEHeader{"Content-Type": {p.contentType}})
+		if err != nil {
+			return "", fmt.Errorf("building vendor-data part: %w", err)
+		}
+		if _, err := io.WriteString(part, p.content); err != nil {
+			return "", fmt.Errorf("writing vendor-data part: %w", err)
+		}
+	}
+	if err := w.Close(); err != nil {
+		return "", fmt.Errorf("closing vendor-data: %w", err)
+	}
+	return fmt.Sprintf("MIME-Version: 1.0\r\nContent-Type: multipart/mixed; boundary=%q\r\n\r\n%s",
+		w.Boundary(), body.String()), nil
+}
+
+// buildSeedISO renders a NoCloud seed: user-data and vendor-data verbatim,
+// and a meta-data whose instance-id is the deterministic VM name, so
 // cloud-init never mistakes a reboot for a new instance. The ISO's volume
 // descriptor carries the build time, so two builds differ in a few
 // timestamp bytes; the files and label are identical.

internal/provider/kvm/seed_test.go

@@ -4,6 +4,8 @@ import (
 	"bytes"
 	"io"
 	"maps"
+	"mime"
+	"mime/multipart"
 	"strings"
 	"testing"
 
@@ -42,7 +44,7 @@ func readSeed(t *testing.T, iso []byte) (label string, files map[string]string)
 func TestBuildSeedISO_contents(t *testing.T) {
 	t.Parallel()
 	userData := "#cloud-config\npackages: [squid]\n"
-	iso, err := buildSeedISO("proxy-abc123def456ghij", userData, defaultVendorData)
+	iso, err := buildSeedISO("proxy-abc123def456ghij", userData, guestAgentScript)
 	if err != nil {
 		t.Fatalf("buildSeedISO() error = %v", err)
 	}
@@ -56,7 +58,7 @@ func TestBuildSeedISO_contents(t *testing.T) {
 	if files["user-data"] != userData {
 		t.Errorf("user-data = %q, want verbatim cloud-init", files["user-data"])
 	}
-	if files["vendor-data"] != defaultVendorData {
+	if files["vendor-data"] != guestAgentScript {
 		t.Errorf("vendor-data = %q", files["vendor-data"])
 	}
 	if got, want := files["meta-data"], "instance-id: proxy-abc123def456ghij\nlocal-hostname: proxy-abc123def456ghij\n"; got != want {
@@ -89,7 +91,7 @@ func TestBuildSeedISO_deterministicContent(t *testing.T) {
 
 func TestBuildSeedISO_emptyUserData(t *testing.T) {
 	t.Parallel()
-	iso, err := buildSeedISO("proxy-x", "", defaultVendorData)
+	iso, err := buildSeedISO("proxy-x", "", guestAgentScript)
 	if err != nil {
 		t.Fatalf("buildSeedISO() error = %v", err)
 	}
@@ -98,3 +100,134 @@ func TestBuildSeedISO_emptyUserData(t *testing.T) {
 		t.Errorf("user-data = %q (present=%v), want an empty file to still exist", body, ok)
 	}
 }
+
+// The whole point of shipping the agent as vendor-data: whatever the Proxy
+// wrote reaches the guest untouched, in any of cloud-init's formats. The
+// previous implementation parsed and re-rendered cloud-config user-data,
+// which dropped comments, reordered keys and re-typed unquoted scalars
+// (version: 1.10 arrived as 1.1).
+func TestBuildSeedISO_userDataIsNeverRewritten(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name     string
+		userData string
+	}{
+		{"cloud-config with its own packages", "#cloud-config\n# keep me\npackages:\n  - squid\nversion: 1.10\nchpasswd: {list: 'u:0123456789'}\n"},
+		{"jinja template", "## template: jinja\n#cloud-config\nhostname: {{ v1.local_hostname }}\n"},
+		{"shell script", "#!/bin/sh\napt-get install -y squid\n"},
+		{"include", "#include\nhttps://example.invalid/x\n"},
+		{"mime multipart", "MIME-Version: 1.0\r\nContent-Type: multipart/mixed; boundary=\"x\"\r\n\r\n--x\r\nContent-Type: text/cloud-config\r\n\r\n#cloud-config\npackages: [squid]\n\r\n--x--\r\n"},
+		{"gzip payload", "\x1f\x8b\x08\x00\x00\x00\x00\x00\x00\x03binary\x00bytes"},
+		{"leading blank line", "\n\n#cloud-config\npackages: [squid]\n"},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			vendor, err := buildVendorData("", true)
+			if err != nil {
+				t.Fatalf("buildVendorData() error = %v", err)
+			}
+			iso, err := buildSeedISO("proxy-x", tc.userData, vendor)
+			if err != nil {
+				t.Fatalf("buildSeedISO() error = %v", err)
+			}
+			_, files := readSeed(t, iso)
+			if files["user-data"] != tc.userData {
+				t.Errorf("user-data was rewritten:\n got %q\nwant %q", files["user-data"], tc.userData)
+			}
+			if !strings.Contains(files["vendor-data"], "qemu-guest-agent") {
+				t.Errorf("vendor-data = %q, want the agent script", files["vendor-data"])
+			}
+		})
+	}
+}
+
+func TestBuildVendorData_agentScriptAlone(t *testing.T) {
+	t.Parallel()
+	got, err := buildVendorData("", true)
+	if err != nil {
+		t.Fatalf("buildVendorData() error = %v", err)
+	}
+	if got != guestAgentScript {
+		t.Errorf("vendor-data = %q, want the bare script", got)
+	}
+	if !strings.HasPrefix(got, "#!") {
+		t.Error("vendor-data must start with a shebang or cloud-init treats it as something other than a script")
+	}
+}
+
+func TestBuildVendorData_optOutKeepsConfiguredOnly(t *testing.T) {
+	t.Parallel()
+	configured := "#cloud-config\nntp: {enabled: true}\n"
+	tests := []struct {
+		name       string
+		configured string
+		want       string
+	}{
+		{"nothing configured", "", ""},
+		{"configured document passes through", configured, configured},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			got, err := buildVendorData(tc.configured, false)
+			if err != nil {
+				t.Fatalf("buildVendorData() error = %v", err)
+			}
+			if got != tc.want {
+				t.Errorf("vendor-data = %q, want %q", got, tc.want)
+			}
+			if strings.Contains(got, "qemu-guest-agent") {
+				t.Error("agent script present although installGuestAgent is off")
+			}
+		})
+	}
+}
+
+// Both documents have to survive in one file, and cloud-init only looks for
+// the multipart marker in the first 4096 bytes.
+func TestBuildVendorData_multipartCarriesBoth(t *testing.T) {
+	t.Parallel()
+	configured := "#cloud-config\nntp:\n  enabled: true\n"
+	got, err := buildVendorData(configured, true)
+	if err != nil {
+		t.Fatalf("buildVendorData() error = %v", err)
+	}
+	head := got
+	if len(head) > 4096 {
+		head = head[:4096]
+	}
+	if !strings.Contains(strings.ToLower(head), "mime-version:") {
+		t.Fatalf("no mime-version: in the first 4096 bytes; cloud-init would treat this as a single document:\n%s", head)
+	}
+	_, params, err := mime.ParseMediaType(strings.SplitN(got, "\r\n\r\n", 2)[0][strings.Index(got, "Content-Type:")+len("Content-Type: "):])
+	if err != nil {
+		t.Fatalf("parsing the multipart content type: %v", err)
+	}
+	body := strings.SplitN(got, "\r\n\r\n", 2)[1]
+	r := multipart.NewReader(strings.NewReader(body), params["boundary"])
+	var bodies, types []string
+	for {
+		part, err := r.NextPart()
+		if err == io.EOF {
+			break
+		}
+		if err != nil {
+			t.Fatalf("reading part: %v", err)
+		}
+		content, err := io.ReadAll(part)
+		if err != nil {
+			t.Fatalf("reading part body: %v", err)
+		}
+		bodies = append(bodies, string(content))
+		types = append(types, part.Header.Get("Content-Type"))
+	}
+	if len(bodies) != 2 || bodies[0] != configured || bodies[1] != guestAgentScript {
+		t.Fatalf("parts = %q, want the configured document then the script", bodies)
+	}
+	// text/plain makes cloud-init re-sniff the user's document from its own
+	// header line, so this code never has to classify it.
+	if types[0] != "text/plain" || types[1] != "text/x-shellscript" {
+		t.Errorf("part types = %v, want [text/plain text/x-shellscript]", types)
+	}
+}

internal/provider/kvm/trace.go

@@ -123,7 +123,10 @@ func (t *tracingAPI) DomainGetMetadata(d libvirt.Domain, typ int32, uri libvirt.
 func (t *tracingAPI) DomainInterfaceAddresses(d libvirt.Domain, source uint32, flags uint32) ([]libvirt.DomainInterface, error) {
 	c := t.begin("DomainInterfaceAddresses", "domain", d.Name, "source", libvirt.DomainInterfaceAddressesSource(source))
 	ifaces, err := t.inner.DomainInterfaceAddresses(d, source, flags)
-	c.detail("addresses", formatAddresses(ifaces)).done(err, "interfaces", len(ifaces))
+	if t.log.V(2).Enabled() {
+		c.detail("addresses", formatAddresses(ifaces))
+	}
+	c.done(err, "interfaces", len(ifaces))
 	return ifaces, err
 }
 
@@ -140,11 +143,14 @@ func formatAddresses(ifaces []libvirt.DomainInterface) []string {
 func (t *tracingAPI) ConnectListAllDomains(needResults int32, flags libvirt.ConnectListAllDomainsFlags) ([]libvirt.Domain, uint32, error) {
 	c := t.begin("ConnectListAllDomains", flagKV(int64(flags))...)
 	doms, n, err := t.inner.ConnectListAllDomains(needResults, flags)
+	if t.log.V(2).Enabled() {
 		names := make([]string, 0, len(doms))
 		for _, d := range doms {
 			names = append(names, d.Name)
 		}
-	c.detail("names", names).done(err, "count", len(doms))
+		c.detail("names", names)
+	}
+	c.done(err, "count", len(doms))
 	return doms, n, err
 }
 
@@ -203,3 +209,24 @@ func (t *tracingAPI) StorageVolDelete(vol libvirt.StorageVol, flags libvirt.Stor
 	c.done(err)
 	return err
 }
+
+func (t *tracingAPI) ConnectGetLibVersion() (uint64, error) {
+	c := t.begin("ConnectGetLibVersion")
+	v, err := t.inner.ConnectGetLibVersion()
+	c.done(err, "version", formatLibVersion(v))
+	return v, err
+}
+
+func (t *tracingAPI) ConnectGetHostname() (string, error) {
+	c := t.begin("ConnectGetHostname")
+	h, err := t.inner.ConnectGetHostname()
+	c.done(err, "hostname", h)
+	return h, err
+}
+
+func (t *tracingAPI) StoragePoolGetInfo(pool libvirt.StoragePool) (uint8, uint64, uint64, uint64, error) {
+	c := t.begin("StoragePoolGetInfo", "pool", pool.Name)
+	state, capacity, allocation, available, err := t.inner.StoragePoolGetInfo(pool)
+	c.done(err, "state", poolStateName(libvirt.StoragePoolState(state)), "capacity", capacity, "available", available)
+	return state, capacity, allocation, available, err
+}

internal/provider/kvm/trace_test.go

@@ -103,6 +103,9 @@ func TestTracingAPI_logsEveryMethod(t *testing.T) {
 			return api.StorageVolDownload(overlay, io.Discard, 0, 0, 0)
 		}},
 		{"StorageVolDelete", func(api libvirtAPI) error { return api.StorageVolDelete(overlay, 0) }},
+		{"ConnectGetLibVersion", func(api libvirtAPI) error { _, err := api.ConnectGetLibVersion(); return err }},
+		{"ConnectGetHostname", func(api libvirtAPI) error { _, err := api.ConnectGetHostname(); return err }},
+		{"StoragePoolGetInfo", func(api libvirtAPI) error { _, _, _, _, err := api.StoragePoolGetInfo(pool); return err }},
 	}
 	if n := reflect.TypeOf((*libvirtAPI)(nil)).Elem().NumMethod(); n != len(cases) {
 		t.Fatalf("libvirtAPI has %d methods, this test covers %d; trace the new one", n, len(cases))

internal/providerhealth/envtest_test.go

@@ -0,0 +1,100 @@
+package providerhealth
+
+import (
+	"context"
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+	"time"
+	"unicode/utf8"
+
+	apierrors "k8s.io/apimachinery/pkg/api/errors"
+	"k8s.io/apimachinery/pkg/runtime"
+	"sigs.k8s.io/controller-runtime/pkg/client"
+	"sigs.k8s.io/controller-runtime/pkg/envtest"
+
+	crawlv1alpha1 "gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/api/v1alpha1"
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+// TestMonitor_envtestRoundTrip writes through a real API server, which is
+// the only thing that validates the generated schema: cluster scope, the
+// status subresource, required fields, list-map keys and the message limit.
+func TestMonitor_envtestRoundTrip(t *testing.T) {
+	if testing.Short() {
+		t.Skip("envtest skipped under -short")
+	}
+	env := &envtest.Environment{
+		CRDDirectoryPaths:     []string{filepath.Join("..", "..", "config", "crd", "bases")},
+		ErrorIfCRDPathMissing: true,
+		BinaryAssetsDirectory: envtestBinaryDir(),
+	}
+	cfg, err := env.Start()
+	if err != nil {
+		t.Fatalf("starting envtest (run `make setup-envtest` first): %v", err)
+	}
+	t.Cleanup(func() { _ = env.Stop() })
+	scheme := runtime.NewScheme()
+	if err := crawlv1alpha1.AddToScheme(scheme); err != nil {
+		t.Fatal(err)
+	}
+	c, err := client.New(cfg, client.Options{Scheme: scheme})
+	if err != nil {
+		t.Fatal(err)
+	}
+	ctx := context.Background()
+	m := &Monitor{Client: c, Reader: c}
+
+	first := Result{Provider: "kvm-home", Type: "kvm", Supported: true, CheckedAt: t0, Targets: []provider.TargetStatus{
+		{Name: "nas1", Address: "nas1:22", Ready: true, Reason: provider.ReasonReachable,
+			Latency: 42 * time.Millisecond, Details: map[string]string{"libvirtVersion": "10.2.0"}},
+		{Name: "nas2", Address: "nas2:22", Reason: provider.ReasonUnreachable, Message: strings.Repeat("x", 2000)},
+	}}
+	if err := m.publish(ctx, first); err != nil {
+		t.Fatalf("first publish: %v", err)
+	}
+	got := getHealth(t, c, "kvm-home")
+	if got.Status.TargetsReady != "1/2" || got.Status.Targets[0].Details["libvirtVersion"] != "10.2.0" ||
+		got.Status.Targets[0].LatencyMillis != 42 || utf8.RuneCountInString(got.Status.Targets[1].Message) != maxMessageLen {
+		t.Errorf("status after the first round = %+v", got.Status)
+	}
+
+	second := Result{Provider: "kvm-home", Type: "kvm", Supported: true, CheckedAt: t1,
+		Targets: []provider.TargetStatus{up("nas1"), up("nas2")}}
+	if err := m.publish(ctx, second); err != nil {
+		t.Fatalf("second publish: %v", err)
+	}
+	got = getHealth(t, c, "kvm-home")
+	if cond := readyCondition(got); cond == nil || cond.Reason != reasonAllTargetsReady {
+		t.Errorf("Ready after recovery = %+v, want %s", cond, reasonAllTargetsReady)
+	}
+	if lt := got.Status.Targets[0].LastTransitionTime.Time; !lt.Equal(t0) {
+		t.Errorf("nas1 lastTransitionTime = %s, want %s", lt, t0)
+	}
+	if lt := got.Status.Targets[1].LastTransitionTime.Time; !lt.Equal(t1) {
+		t.Errorf("nas2 lastTransitionTime = %s, want %s", lt, t1)
+	}
+
+	m.cleanup(ctx, nil)
+	var gone crawlv1alpha1.ProviderHealth
+	if err := c.Get(ctx, client.ObjectKey{Name: "kvm-home"}, &gone); !apierrors.IsNotFound(err) {
+		t.Errorf("Get after cleanup = %v, want NotFound", err)
+	}
+}
+
+// envtestBinaryDir finds the binaries `make setup-envtest` downloads, so
+// the test also runs outside make (which sets KUBEBUILDER_ASSETS instead).
+func envtestBinaryDir() string {
+	base := filepath.Join("..", "..", "bin", "k8s")
+	entries, err := os.ReadDir(base)
+	if err != nil {
+		return ""
+	}
+	for _, e := range entries {
+		if e.IsDir() {
+			return filepath.Join(base, e.Name())
+		}
+	}
+	return ""
+}

internal/providerhealth/monitor.go

@@ -0,0 +1,296 @@
+// Package providerhealth checks every configured provider on an interval
+// and publishes the results: one ProviderHealth object per provider, a log
+// line whenever a target goes down or comes back, and a snapshot the
+// metrics collector reads at scrape time.
+package providerhealth
+
+import (
+	"context"
+	"fmt"
+	"slices"
+	"sync"
+	"time"
+
+	"github.com/go-logr/logr"
+	"go.opentelemetry.io/otel/attribute"
+	"k8s.io/apimachinery/pkg/api/equality"
+	apierrors "k8s.io/apimachinery/pkg/api/errors"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/util/validation"
+	"sigs.k8s.io/controller-runtime/pkg/client"
+	logf "sigs.k8s.io/controller-runtime/pkg/log"
+
+	crawlv1alpha1 "gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/api/v1alpha1"
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/tracing"
+)
+
+const (
+	// labelManagedBy marks the ProviderHealth objects this operator owns;
+	// cleanup never deletes an object without it.
+	labelManagedBy = "app.kubernetes.io/managed-by"
+	managedByValue = "egress-proxies-operator"
+
+	minInterval    = 10 * time.Second
+	defaultTimeout = 15 * time.Second
+)
+
+// Target is one configured provider. Checker is nil for a provider that
+// cannot check itself.
+type Target struct {
+	Name    string
+	Type    string
+	Checker provider.Checker
+}
+
+// Result is the outcome of one provider's latest check.
+type Result struct {
+	Provider string
+	Type     string
+	// Supported is false when the provider has no Checker.
+	Supported bool
+	Targets   []provider.TargetStatus
+	CheckedAt time.Time
+}
+
+// controller-gen reads RBAC markers only at package level, so they stay
+// out of Monitor's doc comment.
+// +kubebuilder:rbac:groups=crawl.example.com,resources=providerhealths,verbs=get;list;create;delete
+// +kubebuilder:rbac:groups=crawl.example.com,resources=providerhealths/status,verbs=get;patch
+
+// Monitor is the manager Runnable that checks the providers.
+type Monitor struct {
+	// Client writes the ProviderHealth objects.
+	Client client.Client
+	// Reader reads them; pass the manager's API reader. The objects are
+	// cluster-scoped, and reading them through the cache would start a
+	// cluster-wide informer needing watch rights, whatever --proxy-namespace
+	// says, for a handful of reads a minute.
+	Reader client.Reader
+	// Targets are the configured providers, in config order.
+	Targets []Target
+	// Interval between rounds of checks; 0 turns checking off. The first
+	// round runs at start: nothing races it, and early visibility is the
+	// point.
+	Interval time.Duration
+	// Timeout bounds each round; 0 means 15s, or the interval when that is
+	// shorter.
+	Timeout time.Duration
+
+	now func() time.Time
+
+	mu     sync.Mutex
+	latest []Result
+
+	// prev holds each target's previous verdict for transition logging.
+	// Only the loop goroutine touches it.
+	prev map[string]provider.TargetStatus
+}
+
+// NeedLeaderElection is true: only the leader dials the hypervisors and
+// writes the objects, and only its gauges describe the checks.
+func (m *Monitor) NeedLeaderElection() bool { return true }
+
+// ValidateSchedule checks --provider-check-interval and
+// --provider-check-timeout. The interval is 0 (off) or at least 10s, which
+// is what keeps one status write per provider per round cheap. The timeout
+// is 0 (derived from the interval, see Monitor.Timeout) or positive and no
+// longer than the interval.
+func ValidateSchedule(interval, timeout time.Duration) error {
+	switch {
+	case interval == 0:
+		return nil
+	case interval < minInterval:
+		return fmt.Errorf("--provider-check-interval must be 0 (off) or at least %s, got %s", minInterval, interval)
+	case timeout < 0 || timeout > interval:
+		return fmt.Errorf("--provider-check-timeout must be at most the interval (%s), got %s", interval, timeout)
+	}
+	return nil
+}
+
+func (m *Monitor) setDefaults() {
+	if m.Timeout == 0 {
+		m.Timeout = min(defaultTimeout, m.Interval)
+	}
+	if m.now == nil {
+		m.now = time.Now
+	}
+}
+
+// Start deletes the objects of providers that are no longer configured,
+// then checks every Interval until ctx ends. The providers config is read
+// once per process, so cleaning up once is enough. With checking off,
+// every object goes, so none is left claiming a stale Ready.
+func (m *Monitor) Start(ctx context.Context) error {
+	m.setDefaults()
+	log := logf.FromContext(ctx).WithName("provider-health")
+	ctx = logf.IntoContext(ctx, log)
+
+	keep := map[string]bool{}
+	if m.Interval > 0 {
+		for _, t := range m.Targets {
+			keep[t.Name] = true
+		}
+	}
+	m.cleanup(ctx, keep)
+	if m.Interval == 0 {
+		log.Info("Provider checks are off (--provider-check-interval=0)")
+		return nil
+	}
+	for _, t := range m.Targets {
+		if errs := validation.IsDNS1123Subdomain(t.Name); len(errs) > 0 {
+			log.Info("WARNING: provider name is not a valid object name; its ProviderHealth object is not written, its metrics are",
+				"provider", t.Name, "problems", errs)
+		}
+	}
+
+	m.checkOnce(ctx)
+	ticker := time.NewTicker(m.Interval)
+	defer ticker.Stop()
+	for {
+		select {
+		case <-ctx.Done():
+			return nil
+		case <-ticker.C:
+			m.checkOnce(ctx)
+		}
+	}
+}
+
+// Snapshot returns the latest results in config order.
+func (m *Monitor) Snapshot() []Result {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	out := make([]Result, len(m.latest))
+	for i, r := range m.latest {
+		r.Targets = slices.Clone(r.Targets)
+		out[i] = r
+	}
+	return out
+}
+
+// checkOnce runs every provider's check concurrently under one timeout,
+// then records, logs and publishes the results provider by provider. A
+// failed write is logged and retried by the next round.
+func (m *Monitor) checkOnce(ctx context.Context) {
+	ctx, span := tracing.Start(ctx, "providerhealth.check")
+	defer span.End()
+	log := logf.FromContext(ctx)
+
+	checkCtx, cancel := context.WithTimeout(ctx, m.Timeout)
+	defer cancel()
+	results := make([]Result, len(m.Targets))
+	var wg sync.WaitGroup
+	for i, t := range m.Targets {
+		wg.Go(func() { results[i] = m.run(checkCtx, t) })
+	}
+	wg.Wait()
+
+	m.mu.Lock()
+	m.latest = results
+	m.mu.Unlock()
+
+	notReady := 0
+	for _, r := range results {
+		m.logTransitions(log, r)
+		for _, ts := range r.Targets {
+			if !ts.Ready {
+				notReady++
+			}
+		}
+		if err := m.publish(ctx, r); err != nil {
+			log.Error(err, "Writing provider health", "provider", r.Provider)
+		}
+	}
+	span.SetAttributes(
+		attribute.Int("providerhealth.providers", len(results)),
+		attribute.Int("providerhealth.targets_not_ready", notReady),
+	)
+}
+
+func (m *Monitor) run(ctx context.Context, t Target) Result {
+	r := Result{Provider: t.Name, Type: t.Type, Supported: t.Checker != nil}
+	if t.Checker != nil {
+		r.Targets = t.Checker.Check(ctx)
+	}
+	r.CheckedAt = m.now()
+	return r
+}
+
+// logTransitions logs a target going down, staying down for a different
+// reason, or coming back. Steady states stay quiet; a target that is ready
+// the first time it is seen only logs at V(1).
+func (m *Monitor) logTransitions(log logr.Logger, r Result) {
+	if m.prev == nil {
+		m.prev = map[string]provider.TargetStatus{}
+	}
+	for _, ts := range r.Targets {
+		key := r.Provider + "/" + ts.Name
+		prev, seen := m.prev[key]
+		m.prev[key] = ts
+		kv := []any{"provider", r.Provider, "target", ts.Name, "address", ts.Address, "reason", ts.Reason}
+		switch {
+		case !ts.Ready && (!seen || prev.Ready || prev.Reason != ts.Reason):
+			log.Info("WARNING: provider target not ready", append(kv, "message", ts.Message)...)
+		case ts.Ready && seen && !prev.Ready:
+			log.Info("Provider target ready", append(kv, "latency", ts.Latency.Round(time.Millisecond))...)
+		case ts.Ready && !seen:
+			log.V(1).Info("Provider target ready", append(kv, "latency", ts.Latency.Round(time.Millisecond))...)
+		}
+	}
+}
+
+// publish creates the provider's ProviderHealth object when missing and
+// patches its status. The status changes on every round, since
+// lastCheckTime moves; the comparison only skips the patch when a clock
+// did not advance.
+func (m *Monitor) publish(ctx context.Context, r Result) error {
+	if len(validation.IsDNS1123Subdomain(r.Provider)) > 0 {
+		return nil
+	}
+	var ph crawlv1alpha1.ProviderHealth
+	err := m.Reader.Get(ctx, client.ObjectKey{Name: r.Provider}, &ph)
+	switch {
+	case apierrors.IsNotFound(err):
+		ph = crawlv1alpha1.ProviderHealth{ObjectMeta: metav1.ObjectMeta{
+			Name:   r.Provider,
+			Labels: map[string]string{labelManagedBy: managedByValue},
+		}}
+		if err := m.Client.Create(ctx, &ph); err != nil {
+			return fmt.Errorf("creating ProviderHealth %s: %w", r.Provider, err)
+		}
+	case err != nil:
+		return fmt.Errorf("reading ProviderHealth %s: %w", r.Provider, err)
+	}
+	base := ph.DeepCopy()
+	ph.Status = buildStatus(ph.Status, r)
+	if equality.Semantic.DeepEqual(base.Status, ph.Status) {
+		return nil
+	}
+	if err := m.Client.Status().Patch(ctx, &ph, client.MergeFrom(base)); err != nil {
+		return fmt.Errorf("patching ProviderHealth %s status: %w", r.Provider, err)
+	}
+	return nil
+}
+
+// cleanup deletes this operator's ProviderHealth objects whose name is not
+// in keep.
+func (m *Monitor) cleanup(ctx context.Context, keep map[string]bool) {
+	log := logf.FromContext(ctx)
+	var list crawlv1alpha1.ProviderHealthList
+	if err := m.Reader.List(ctx, &list, client.MatchingLabels{labelManagedBy: managedByValue}); err != nil {
+		log.Error(err, "Listing provider health objects; stale ones stay until the next start")
+		return
+	}
+	for i := range list.Items {
+		ph := &list.Items[i]
+		if keep[ph.Name] {
+			continue
+		}
+		if err := m.Client.Delete(ctx, ph); client.IgnoreNotFound(err) != nil {
+			log.Error(err, "Deleting provider health of an unconfigured provider", "name", ph.Name)
+			continue
+		}
+		log.Info("Deleted provider health of an unconfigured provider", "name", ph.Name)
+	}
+}

internal/providerhealth/monitor_test.go

@@ -0,0 +1,377 @@
+package providerhealth
+
+import (
+	"context"
+	"strings"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/go-logr/logr"
+	"github.com/go-logr/logr/funcr"
+	apierrors "k8s.io/apimachinery/pkg/api/errors"
+	apimeta "k8s.io/apimachinery/pkg/api/meta"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/runtime"
+	"k8s.io/apimachinery/pkg/runtime/schema"
+	"sigs.k8s.io/controller-runtime/pkg/client"
+	"sigs.k8s.io/controller-runtime/pkg/client/fake"
+
+	crawlv1alpha1 "gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/api/v1alpha1"
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+var (
+	t0 = time.Date(2026, 9, 11, 15, 0, 0, 0, time.UTC)
+	t1 = t0.Add(time.Minute)
+)
+
+type checkFunc func(ctx context.Context) []provider.TargetStatus
+
+func (f checkFunc) Check(ctx context.Context) []provider.TargetStatus { return f(ctx) }
+
+func verdicts(ts ...provider.TargetStatus) checkFunc {
+	return func(context.Context) []provider.TargetStatus { return ts }
+}
+
+func up(name string) provider.TargetStatus {
+	return provider.TargetStatus{Name: name, Address: name + ":22", Ready: true, Reason: provider.ReasonReachable}
+}
+
+func down(name, reason string) provider.TargetStatus {
+	return provider.TargetStatus{Name: name, Address: name + ":22", Reason: reason, Message: name + " is down"}
+}
+
+func managed(name string) *crawlv1alpha1.ProviderHealth {
+	return &crawlv1alpha1.ProviderHealth{ObjectMeta: metav1.ObjectMeta{
+		Name:   name,
+		Labels: map[string]string{labelManagedBy: managedByValue},
+	}}
+}
+
+// newFakeClient registers ProviderHealth as cluster-scoped; the fake's
+// default mapper would treat it as namespaced.
+func newFakeClient(t *testing.T, objs ...client.Object) client.WithWatch {
+	t.Helper()
+	scheme := runtime.NewScheme()
+	if err := crawlv1alpha1.AddToScheme(scheme); err != nil {
+		t.Fatal(err)
+	}
+	mapper := apimeta.NewDefaultRESTMapper([]schema.GroupVersion{crawlv1alpha1.GroupVersion})
+	mapper.Add(crawlv1alpha1.GroupVersion.WithKind("ProviderHealth"), apimeta.RESTScopeRoot)
+	return fake.NewClientBuilder().
+		WithScheme(scheme).
+		WithRESTMapper(mapper).
+		WithStatusSubresource(&crawlv1alpha1.ProviderHealth{}).
+		WithObjects(objs...).
+		Build()
+}
+
+func newTestMonitor(c client.Client, targets ...Target) *Monitor {
+	return &Monitor{
+		Client:   c,
+		Reader:   c,
+		Targets:  targets,
+		Interval: time.Minute,
+		Timeout:  time.Second,
+		now:      func() time.Time { return t1 },
+	}
+}
+
+func getHealth(t *testing.T, c client.Reader, name string) *crawlv1alpha1.ProviderHealth {
+	t.Helper()
+	var ph crawlv1alpha1.ProviderHealth
+	if err := c.Get(context.Background(), client.ObjectKey{Name: name}, &ph); err != nil {
+		t.Fatalf("Get ProviderHealth %s: %v", name, err)
+	}
+	return &ph
+}
+
+func readyCondition(ph *crawlv1alpha1.ProviderHealth) *metav1.Condition {
+	return apimeta.FindStatusCondition(ph.Status.Conditions, crawlv1alpha1.ConditionReady)
+}
+
+func TestMonitor_checkOnce_createsLabelledObjects(t *testing.T) {
+	t.Parallel()
+	c := newFakeClient(t)
+	m := newTestMonitor(c,
+		Target{Name: "kvm-home", Type: "kvm", Checker: verdicts(up("nas1"), down("nas2", provider.ReasonUnreachable))},
+		Target{Name: "local", Type: "kubernetes", Checker: verdicts(up("apiserver"))},
+	)
+
+	m.checkOnce(context.Background())
+
+	kvm := getHealth(t, c, "kvm-home")
+	if kvm.Labels[labelManagedBy] != managedByValue {
+		t.Errorf("labels = %v, want %s=%s", kvm.Labels, labelManagedBy, managedByValue)
+	}
+	if cond := readyCondition(kvm); cond == nil || cond.Status != metav1.ConditionFalse || cond.Reason != reasonTargetsNotReady {
+		t.Errorf("kvm-home Ready = %+v, want False/%s", cond, reasonTargetsNotReady)
+	}
+	if kvm.Status.Type != "kvm" || kvm.Status.TargetsReady != "1/2" || kvm.Status.LastCheckTime == nil || !kvm.Status.LastCheckTime.Time.Equal(t1) {
+		t.Errorf("kvm-home status = %+v, want type kvm, 1/2 ready, checked at %s", kvm.Status, t1)
+	}
+	local := getHealth(t, c, "local")
+	if cond := readyCondition(local); cond == nil || cond.Status != metav1.ConditionTrue {
+		t.Errorf("local Ready = %+v, want True", cond)
+	}
+}
+
+func TestMonitor_checkOnce_patchesExisting(t *testing.T) {
+	t.Parallel()
+	existing := managed("local")
+	existing.Status = crawlv1alpha1.ProviderHealthStatus{
+		Type:    "kubernetes",
+		Targets: []crawlv1alpha1.ProviderTarget{readyAt("apiserver", t0)},
+	}
+	c := newFakeClient(t, existing)
+	m := newTestMonitor(c, Target{Name: "local", Type: "kubernetes", Checker: verdicts(up("apiserver"))})
+
+	m.checkOnce(context.Background())
+
+	got := getHealth(t, c, "local")
+	if got.Status.LastCheckTime == nil || !got.Status.LastCheckTime.Time.Equal(t1) {
+		t.Errorf("lastCheckTime = %v, want %s", got.Status.LastCheckTime, t1)
+	}
+	if lt := got.Status.Targets[0].LastTransitionTime.Time; !lt.Equal(t0) {
+		t.Errorf("apiserver lastTransitionTime = %s, want %s kept: it was ready before too", lt, t0)
+	}
+}
+
+func TestMonitor_checkOnce_timeoutBoundsEachRound(t *testing.T) {
+	t.Parallel()
+	c := newFakeClient(t)
+	var hadDeadline atomic.Bool
+	hanging := checkFunc(func(ctx context.Context) []provider.TargetStatus {
+		_, ok := ctx.Deadline()
+		hadDeadline.Store(ok)
+		<-ctx.Done()
+		return []provider.TargetStatus{down("nas1", provider.ReasonTimeout)}
+	})
+	m := newTestMonitor(c,
+		Target{Name: "kvm-home", Type: "kvm", Checker: hanging},
+		Target{Name: "local", Type: "kubernetes", Checker: verdicts(up("apiserver"))},
+	)
+	m.Timeout = 50 * time.Millisecond
+
+	done := make(chan struct{})
+	go func() {
+		m.checkOnce(context.Background())
+		close(done)
+	}()
+	select {
+	case <-done:
+	case <-time.After(5 * time.Second):
+		t.Fatal("checkOnce still running 5s after a 50ms timeout")
+	}
+
+	if !hadDeadline.Load() {
+		t.Error("the checker got a context without a deadline")
+	}
+	if cond := readyCondition(getHealth(t, c, "kvm-home")); cond == nil || cond.Status != metav1.ConditionFalse {
+		t.Errorf("kvm-home Ready = %+v, want False", cond)
+	}
+	if cond := readyCondition(getHealth(t, c, "local")); cond == nil || cond.Status != metav1.ConditionTrue {
+		t.Errorf("local Ready = %+v, want True despite the slow neighbour", cond)
+	}
+}
+
+func TestMonitor_checkOnce_noCheckerIsUnknown(t *testing.T) {
+	t.Parallel()
+	c := newFakeClient(t)
+	m := newTestMonitor(c, Target{Name: "gcp-eu", Type: "gcp"})
+
+	m.checkOnce(context.Background())
+
+	got := getHealth(t, c, "gcp-eu")
+	if cond := readyCondition(got); cond == nil || cond.Status != metav1.ConditionUnknown || cond.Reason != reasonChecksNotImplemented {
+		t.Errorf("Ready = %+v, want Unknown/%s", cond, reasonChecksNotImplemented)
+	}
+}
+
+func TestMonitor_checkOnce_invalidNameSkipsTheObject(t *testing.T) {
+	t.Parallel()
+	c := newFakeClient(t)
+	m := newTestMonitor(c, Target{Name: "Home_KVM", Type: "kvm", Checker: verdicts(up("nas1"))})
+
+	m.checkOnce(context.Background())
+
+	var list crawlv1alpha1.ProviderHealthList
+	if err := c.List(context.Background(), &list); err != nil {
+		t.Fatal(err)
+	}
+	if len(list.Items) != 0 {
+		t.Errorf("wrote %d objects, want none for an invalid name", len(list.Items))
+	}
+	snap := m.Snapshot()
+	if len(snap) != 1 || snap[0].Provider != "Home_KVM" || !snap[0].Targets[0].Ready {
+		t.Errorf("Snapshot() = %+v, want the result kept for metrics", snap)
+	}
+}
+
+func TestMonitor_logsOnlyTransitions(t *testing.T) {
+	t.Parallel()
+	var mu sync.Mutex
+	var lines []string
+	log := funcr.New(func(_, args string) {
+		mu.Lock()
+		defer mu.Unlock()
+		lines = append(lines, args)
+	}, funcr.Options{})
+	ctx := logr.NewContext(context.Background(), log)
+	rounds := [][]provider.TargetStatus{
+		{up("nas1")},
+		{up("nas1")},
+		{down("nas1", provider.ReasonUnreachable)},
+		{down("nas1", provider.ReasonUnreachable)},
+		{down("nas1", provider.ReasonTimeout)},
+		{up("nas1")},
+	}
+	round := 0
+	m := newTestMonitor(newFakeClient(t), Target{Name: "kvm-home", Type: "kvm", Checker: checkFunc(
+		func(context.Context) []provider.TargetStatus { return rounds[round] })})
+
+	for round = range rounds {
+		m.checkOnce(ctx)
+	}
+
+	mu.Lock()
+	defer mu.Unlock()
+	want := [][]string{
+		{`"msg"="WARNING: provider target not ready"`, `"reason"="Unreachable"`},
+		{`"msg"="WARNING: provider target not ready"`, `"reason"="Timeout"`},
+		{`"msg"="Provider target ready"`},
+	}
+	if len(lines) != len(want) {
+		t.Fatalf("logged %d lines, want %d:\n%s", len(lines), len(want), strings.Join(lines, "\n"))
+	}
+	for i, subs := range want {
+		for _, sub := range subs {
+			if !strings.Contains(lines[i], sub) {
+				t.Errorf("line %d missing %s:\n%s", i, sub, lines[i])
+			}
+		}
+	}
+}
+
+func TestMonitor_cleanup_deletesOnlyLabelledUnconfigured(t *testing.T) {
+	t.Parallel()
+	foreign := &crawlv1alpha1.ProviderHealth{ObjectMeta: metav1.ObjectMeta{Name: "foreign"}}
+	c := newFakeClient(t, managed("kvm-home"), managed("retired"), foreign)
+	m := newTestMonitor(c)
+
+	m.cleanup(context.Background(), map[string]bool{"kvm-home": true})
+
+	getHealth(t, c, "kvm-home")
+	getHealth(t, c, "foreign")
+	var gone crawlv1alpha1.ProviderHealth
+	if err := c.Get(context.Background(), client.ObjectKey{Name: "retired"}, &gone); !apierrors.IsNotFound(err) {
+		t.Errorf("Get retired = %v, want NotFound", err)
+	}
+}
+
+func TestMonitor_Start_offCleansUpAndReturns(t *testing.T) {
+	t.Parallel()
+	c := newFakeClient(t, managed("kvm-home"))
+	var checked atomic.Bool
+	m := newTestMonitor(c, Target{Name: "kvm-home", Type: "kvm", Checker: checkFunc(
+		func(context.Context) []provider.TargetStatus { checked.Store(true); return nil })})
+	m.Interval = 0
+
+	errc := make(chan error, 1)
+	go func() { errc <- m.Start(context.Background()) }()
+	select {
+	case err := <-errc:
+		if err != nil {
+			t.Fatalf("Start() = %v, want nil", err)
+		}
+	case <-time.After(5 * time.Second):
+		t.Fatal("Start did not return with checking off")
+	}
+
+	if checked.Load() {
+		t.Error("a check ran with checking off")
+	}
+	var gone crawlv1alpha1.ProviderHealth
+	if err := c.Get(context.Background(), client.ObjectKey{Name: "kvm-home"}, &gone); !apierrors.IsNotFound(err) {
+		t.Errorf("Get kvm-home = %v, want NotFound so no stale Ready is left behind", err)
+	}
+}
+
+func TestMonitor_Start_checksAtOnceAndStops(t *testing.T) {
+	t.Parallel()
+	c := newFakeClient(t)
+	m := newTestMonitor(c, Target{Name: "local", Type: "kubernetes", Checker: verdicts(up("apiserver"))})
+	m.Interval = time.Hour
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+
+	errc := make(chan error, 1)
+	go func() { errc <- m.Start(ctx) }()
+	deadline := time.Now().Add(5 * time.Second)
+	for len(m.Snapshot()) == 0 {
+		if time.Now().After(deadline) {
+			t.Fatal("no check within 5s of Start, want one right away")
+		}
+		time.Sleep(10 * time.Millisecond)
+	}
+	cancel()
+
+	select {
+	case err := <-errc:
+		if err != nil {
+			t.Errorf("Start() = %v, want nil", err)
+		}
+	case <-time.After(5 * time.Second):
+		t.Fatal("Start did not return after its context ended")
+	}
+}
+
+func TestValidateSchedule(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name              string
+		interval, timeout time.Duration
+		wantErr           bool
+	}{
+		{"off", 0, 0, false},
+		{"defaults", time.Minute, 15 * time.Second, false},
+		{"shortest interval", 10 * time.Second, 10 * time.Second, false},
+		{"too frequent", 5 * time.Second, time.Second, true},
+		{"negative interval", -time.Minute, time.Second, true},
+		{"timeout left to derive, short interval", 12 * time.Second, 0, false},
+		{"negative timeout", time.Minute, -time.Second, true},
+		{"timeout longer than interval", time.Minute, 2 * time.Minute, true},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			if err := ValidateSchedule(tc.interval, tc.timeout); (err != nil) != tc.wantErr {
+				t.Errorf("ValidateSchedule(%s, %s) = %v, want error %v", tc.interval, tc.timeout, err, tc.wantErr)
+			}
+		})
+	}
+}
+
+func TestMonitor_setDefaults_timeout(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name                    string
+		interval, timeout, want time.Duration
+	}{
+		{"default", time.Minute, 0, 15 * time.Second},
+		{"interval shorter than the default", 12 * time.Second, 0, 12 * time.Second},
+		{"explicit", time.Minute, 20 * time.Second, 20 * time.Second},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			m := &Monitor{Interval: tc.interval, Timeout: tc.timeout}
+			m.setDefaults()
+			if m.Timeout != tc.want {
+				t.Errorf("Timeout = %s, want %s", m.Timeout, tc.want)
+			}
+		})
+	}
+}

internal/providerhealth/status.go

@@ -0,0 +1,98 @@
+package providerhealth
+
+import (
+	"fmt"
+	"slices"
+	"strings"
+	"unicode/utf8"
+
+	apimeta "k8s.io/apimachinery/pkg/api/meta"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+
+	crawlv1alpha1 "gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/api/v1alpha1"
+)
+
+// Reasons on the Ready condition.
+const (
+	reasonAllTargetsReady      = "AllTargetsReady"
+	reasonTargetsNotReady      = "TargetsNotReady"
+	reasonChecksNotImplemented = "ChecksNotImplemented"
+	reasonNoTargets            = "NoTargets"
+)
+
+// maxMessageLen matches the CRD's limit on a target's message.
+const maxMessageLen = 1024
+
+// buildStatus derives a ProviderHealth status from the previous one and a
+// fresh result. Every timestamp comes from r.CheckedAt, and prev is left
+// untouched. A target keeps its lastTransitionTime while ready does not
+// flip, which is how several hosts each show when they went down.
+func buildStatus(prev crawlv1alpha1.ProviderHealthStatus, r Result) crawlv1alpha1.ProviderHealthStatus {
+	now := metav1.NewTime(r.CheckedAt)
+	out := crawlv1alpha1.ProviderHealthStatus{
+		Type:          r.Type,
+		Conditions:    slices.Clone(prev.Conditions),
+		LastCheckTime: &now,
+	}
+	before := make(map[string]crawlv1alpha1.ProviderTarget, len(prev.Targets))
+	for _, t := range prev.Targets {
+		before[t.Name] = t
+	}
+	ready := 0
+	var failing []string
+	for _, ts := range r.Targets {
+		t := crawlv1alpha1.ProviderTarget{
+			Name:               ts.Name,
+			Address:            ts.Address,
+			Ready:              ts.Ready,
+			Reason:             ts.Reason,
+			Message:            truncate(ts.Message, maxMessageLen),
+			LatencyMillis:      ts.Latency.Milliseconds(),
+			LastTransitionTime: now,
+			Details:            ts.Details,
+		}
+		if b, ok := before[ts.Name]; ok && b.Ready == ts.Ready {
+			t.LastTransitionTime = b.LastTransitionTime
+		}
+		if ts.Ready {
+			ready++
+		} else {
+			failing = append(failing, ts.Name+": "+ts.Reason)
+		}
+		out.Targets = append(out.Targets, t)
+	}
+	if r.Supported {
+		out.TargetsReady = fmt.Sprintf("%d/%d", ready, len(r.Targets))
+	}
+
+	cond := metav1.Condition{Type: crawlv1alpha1.ConditionReady, LastTransitionTime: now}
+	switch {
+	case !r.Supported:
+		cond.Status = metav1.ConditionUnknown
+		cond.Reason = reasonChecksNotImplemented
+		cond.Message = fmt.Sprintf("%s providers have no health check", r.Type)
+	case len(r.Targets) == 0:
+		cond.Status = metav1.ConditionUnknown
+		cond.Reason = reasonNoTargets
+		cond.Message = "the provider has nothing to check"
+	case len(failing) == 0:
+		cond.Status = metav1.ConditionTrue
+		cond.Reason = reasonAllTargetsReady
+		cond.Message = fmt.Sprintf("%d of %d targets ready", ready, len(r.Targets))
+	default:
+		cond.Status = metav1.ConditionFalse
+		cond.Reason = reasonTargetsNotReady
+		cond.Message = truncate(strings.Join(failing, "; "), maxMessageLen)
+	}
+	apimeta.SetStatusCondition(&out.Conditions, cond)
+	return out
+}
+
+// truncate cuts s to at most n characters, which is how the API server
+// counts maxLength.
+func truncate(s string, n int) string {
+	if utf8.RuneCountInString(s) <= n {
+		return s
+	}
+	return string([]rune(s)[:n-1]) + "…"
+}

internal/providerhealth/status_test.go

@@ -0,0 +1,135 @@
+package providerhealth
+
+import (
+	"strings"
+	"testing"
+	"time"
+	"unicode/utf8"
+
+	apimeta "k8s.io/apimachinery/pkg/api/meta"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+
+	crawlv1alpha1 "gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/api/v1alpha1"
+	"gitea.home.hrajfrisbee.cz/kacerr/egress-proxies-operator/internal/provider"
+)
+
+func readyAt(name string, at time.Time) crawlv1alpha1.ProviderTarget {
+	return crawlv1alpha1.ProviderTarget{Name: name, Ready: true, Reason: provider.ReasonReachable, LastTransitionTime: metav1.NewTime(at)}
+}
+
+func TestBuildStatus(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name             string
+		prev             crawlv1alpha1.ProviderHealthStatus
+		result           Result
+		wantStatus       metav1.ConditionStatus
+		wantReason       string
+		wantMessage      string
+		wantTargetsReady string
+		// wantTransitions is each target's lastTransitionTime, in order.
+		wantTransitions []time.Time
+	}{
+		{
+			name:             "all ready",
+			result:           Result{Type: "kvm", Supported: true, Targets: []provider.TargetStatus{up("nas1"), up("nas2")}, CheckedAt: t1},
+			wantStatus:       metav1.ConditionTrue,
+			wantReason:       reasonAllTargetsReady,
+			wantMessage:      "2 of 2 targets ready",
+			wantTargetsReady: "2/2",
+			wantTransitions:  []time.Time{t1, t1},
+		},
+		{
+			name:             "one target down",
+			result:           Result{Type: "kvm", Supported: true, Targets: []provider.TargetStatus{up("nas1"), down("nas2", provider.ReasonUnreachable)}, CheckedAt: t1},
+			wantStatus:       metav1.ConditionFalse,
+			wantReason:       reasonTargetsNotReady,
+			wantMessage:      "nas2: Unreachable",
+			wantTargetsReady: "1/2",
+			wantTransitions:  []time.Time{t1, t1},
+		},
+		{
+			name: "only a flip moves a target's transition time",
+			prev: crawlv1alpha1.ProviderHealthStatus{Targets: []crawlv1alpha1.ProviderTarget{readyAt("nas1", t0), readyAt("nas2", t0)}},
+			result: Result{Type: "kvm", Supported: true, CheckedAt: t1,
+				Targets: []provider.TargetStatus{up("nas1"), down("nas2", provider.ReasonTimeout)}},
+			wantStatus:       metav1.ConditionFalse,
+			wantReason:       reasonTargetsNotReady,
+			wantMessage:      "nas2: Timeout",
+			wantTargetsReady: "1/2",
+			wantTransitions:  []time.Time{t0, t1},
+		},
+		{
+			name:       "no checker",
+			result:     Result{Type: "gcp", CheckedAt: t1},
+			wantStatus: metav1.ConditionUnknown,
+			wantReason: reasonChecksNotImplemented,
+		},
+		{
+			name:             "nothing to check",
+			result:           Result{Type: "kvm", Supported: true, CheckedAt: t1},
+			wantStatus:       metav1.ConditionUnknown,
+			wantReason:       reasonNoTargets,
+			wantTargetsReady: "0/0",
+		},
+	}
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+			got := buildStatus(tc.prev, tc.result)
+
+			cond := apimeta.FindStatusCondition(got.Conditions, crawlv1alpha1.ConditionReady)
+			if cond == nil || cond.Status != tc.wantStatus || cond.Reason != tc.wantReason {
+				t.Fatalf("Ready = %+v, want %s/%s", cond, tc.wantStatus, tc.wantReason)
+			}
+			if tc.wantMessage != "" && cond.Message != tc.wantMessage {
+				t.Errorf("Ready message = %q, want %q", cond.Message, tc.wantMessage)
+			}
+			if got.TargetsReady != tc.wantTargetsReady {
+				t.Errorf("targetsReady = %q, want %q", got.TargetsReady, tc.wantTargetsReady)
+			}
+			if got.Type != tc.result.Type || got.LastCheckTime == nil || !got.LastCheckTime.Time.Equal(t1) {
+				t.Errorf("type/lastCheckTime = %q/%v, want %q/%s", got.Type, got.LastCheckTime, tc.result.Type, t1)
+			}
+			if len(got.Targets) != len(tc.wantTransitions) {
+				t.Fatalf("got %d targets, want %d", len(got.Targets), len(tc.wantTransitions))
+			}
+			for i, want := range tc.wantTransitions {
+				if lt := got.Targets[i].LastTransitionTime.Time; !lt.Equal(want) {
+					t.Errorf("target %s lastTransitionTime = %s, want %s", got.Targets[i].Name, lt, want)
+				}
+			}
+		})
+	}
+}
+
+func TestBuildStatus_conditionKeepsTransitionWhileUnchanged(t *testing.T) {
+	t.Parallel()
+	first := buildStatus(crawlv1alpha1.ProviderHealthStatus{},
+		Result{Type: "kvm", Supported: true, Targets: []provider.TargetStatus{up("nas1")}, CheckedAt: t0})
+	firstCond := apimeta.FindStatusCondition(first.Conditions, crawlv1alpha1.ConditionReady)
+
+	second := buildStatus(first, Result{Type: "kvm", Supported: true, Targets: []provider.TargetStatus{up("nas1")}, CheckedAt: t1})
+
+	cond := apimeta.FindStatusCondition(second.Conditions, crawlv1alpha1.ConditionReady)
+	if !cond.LastTransitionTime.Time.Equal(t0) {
+		t.Errorf("Ready lastTransitionTime = %s, want %s from the first round", cond.LastTransitionTime, t0)
+	}
+	if !firstCond.LastTransitionTime.Time.Equal(t0) || len(first.Conditions) != 1 {
+		t.Errorf("buildStatus modified the previous status: %+v", first.Conditions)
+	}
+}
+
+func TestBuildStatus_truncatesLongMessages(t *testing.T) {
+	t.Parallel()
+	long := strings.Repeat("é", 3000)
+	got := buildStatus(crawlv1alpha1.ProviderHealthStatus{}, Result{Type: "kvm", Supported: true, CheckedAt: t1,
+		Targets: []provider.TargetStatus{{Name: "nas1", Reason: provider.ReasonUnreachable, Message: long}}})
+
+	if n := utf8.RuneCountInString(got.Targets[0].Message); n != maxMessageLen {
+		t.Errorf("message has %d characters, want %d", n, maxMessageLen)
+	}
+	if !strings.HasSuffix(got.Targets[0].Message, "…") {
+		t.Error("truncated message does not end with an ellipsis")
+	}
+}

test/e2e/e2e_test.go

@@ -173,6 +173,20 @@ var _ = Describe("Manager", Ordered, func() {
 			Eventually(verifyControllerUp).Should(Succeed())
 		})
 
+		// Runs before the metrics spec on purpose: once a ProviderHealth is
+		// Ready, the first round of checks is done and its gauges exist.
+		It("should publish provider health", func() {
+			By("waiting for the kubernetes provider's ProviderHealth to report Ready")
+			verifyProviderReady := func(g Gomega) {
+				cmd := exec.Command("kubectl", "get", "providerhealth", "kubernetes",
+					"-o", `jsonpath={.status.conditions[?(@.type=="Ready")].status}`)
+				output, err := utils.Run(cmd)
+				g.Expect(err).NotTo(HaveOccurred())
+				g.Expect(output).To(Equal("True"), "kubernetes provider not Ready")
+			}
+			Eventually(verifyProviderReady).Should(Succeed())
+		})
+
 		It("should ensure the metrics endpoint is serving metrics", func() {
 			By("creating a ClusterRoleBinding for the service account to allow access to metrics")
 			cmd := exec.Command("kubectl", "create", "clusterrolebinding", metricsRoleBindingName,
@@ -264,6 +278,8 @@ var _ = Describe("Manager", Ordered, func() {
 				g.Expect(err).NotTo(HaveOccurred(), "Failed to retrieve logs from curl pod")
 				g.Expect(metricsOutput).NotTo(BeEmpty())
 				g.Expect(metricsOutput).To(ContainSubstring("< HTTP/1.1 200 OK"))
+				g.Expect(metricsOutput).To(ContainSubstring(
+					`proxy_operator_provider_target_up{provider="kubernetes",target="apiserver"} 1`))
 			}
 			Eventually(verifyMetricsAvailable, 2*time.Minute).Should(Succeed())
 		})