# helm upgrade --install ca cluster-autoscaler --repo https://kubernetes.github.io/autoscaler --version 9.59.0 -n default -f values.yaml
# must be installed in the same namespace as the <cluster>-kubeconfig secret
cloudProvider: clusterapi
clusterAPIMode: kubeconfig-incluster      # workload via kubeconfig, management in-cluster
clusterAPIKubeconfigSecret: lab1-kubeconfig
autoDiscovery:
  clusterName: lab1
  namespace: default
image:
  tag: v1.35.2                            # chart default appVersion is 1.35.0
rbac:
  # CA reads the infrastructure machine template (capacity for scale-from-zero). The chart's
  # ClusterRole only covers cluster.x-k8s.io; without this the first main loop blocks forever on
  # "waiting for cache sync on infrastructure resource" and CA never scales.
  additionalRules:
    - apiGroups: [infrastructure.cluster.x-k8s.io]
      resources: [proxmoxmachinetemplates]
      verbs: [get, list, watch]
extraArgs:
  expander: least-waste
  # Aggressive lab scale-down profile (CA 1.35 defaults in brackets): idle worker gone in ~3-4 min
  # (scan-interval 1m adds up to a minute to the 1m unneeded time).
  # Keep delay-after-add >= node startup (~2-3 min here) or bursts flap up/down.
  scale-down-unneeded-time: 1m                # [10m]
  scale-down-delay-after-add: 2m              # [10m] cluster-wide cooldown after any scale-up
  scale-down-delay-after-failure: 1m          # [3m]
  unremovable-node-recheck-timeout: 1m        # [5m]
  scale-down-utilization-threshold: "0.6"     # [0.5]
  max-graceful-termination-sec: "60"          # [600] drain wait per pod
  balance-similar-node-groups: true
  # One worker MD per PVE node. Compare groups only on this label: the generic comparator fails when
  # one group has a real-node template (cached, 2Gi VM = ~1.78Gi allocatable, os/arch set) and the
  # other a from-zero one. Disables the resource checks, so keep the MDs' capacity identical.
  balancing-label: node.cluster.x-k8s.io/pool
  # Fewer API requests instead of a higher kube-client-qps: every MD at 0 replicas costs ~3 uncached
  # discovery GETs per NodeGroups() call (~15x per loop). Unschedulable pods and scale-ups/downs still
  # trigger a loop immediately; only idle loops (scale-down evaluation) are spaced out.
  scan-interval: 1m                           # [10s]
  # Cilium taints new nodes NoSchedule until its agent is up. Without this, CA simulates the freshly
  # registered (still tainted) node, finds the pending pods don't fit and adds one node too many.
  startup-taint: node.cilium.io/agent-not-ready
  v: 2                                        # [chart 4] 4 logs every throttled request and MD lookup