kube-pool-manager

Kube pool manager

Manager for Kubernetes pool, automatic applies configuration (annotations, labels, configSource, role) to kubernetes nodes based on any node spec.

Supports JSON path, value and regexp matching.

Sets following settings on nodes if matched:

• node role
• node labels
• node annotations
• node configSource

Node settings are applied on startup and for new nodes (delayed until they are ready) and (optional) on watch timeout.

Configuration

Usage:
  kube-pool-manager [OPTIONS]

Application Options:
      --log.level=[trace|debug|info|warning|error] Log level (default: info) [$LOG_LEVEL]
      --log.format=[logfmt|json]                   Log format (default: logfmt) [$LOG_FORMAT]
      --log.source=[|short|file|full]              Show source for every log message (useful for debugging and bug reports) [$LOG_SOURCE]
      --log.color=[|auto|yes|no]                   Enable color for logs [$LOG_COLOR]
      --log.time                                   Show log time [$LOG_TIME]
      --instance.nodename=                         Name of node where autopilot is running [$INSTANCE_NODENAME]
      --instance.namespace=                        Name of namespace where autopilot is running [$INSTANCE_NAMESPACE]
      --instance.pod=                              Name of pod where autopilot is running [$INSTANCE_POD]
      --kube.node.labelselector=                   Node Label selector which nodes should be checked [$KUBE_NODE_LABELSELECTOR]
      --kube.watch.timeout=                        Timeout & full resync for node watch (time.Duration) (default: 24h) [$KUBE_WATCH_TIMEOUT]
      --kube.watch.reapply                         Reapply node settings on watch timeout [$KUBE_WATCH_REAPPLY]
      --lease.enable                               Enable lease (leader election; enabled by default in docker images) [$LEASE_ENABLE]
      --lease.name=                                Name of lease lock (default: kube-pool-manager-leader) [$LEASE_NAME]
      --server.bind=                               Server address (default: :8080) [$SERVER_BIND]
      --server.timeout.read=                       Server read timeout (default: 5s) [$SERVER_TIMEOUT_READ]
      --server.timeout.write=                      Server write timeout (default: 10s) [$SERVER_TIMEOUT_WRITE]
      --dry-run                                    Dry run (do not apply to nodes) [$DRY_RUN]
      --config=                                    Config path [$CONFIG]

Help Options:
  -h, --help                                       Show this help message

see example.yaml for configuration file

Metrics

(see :8080/metrics)

Kubernetes deployment

see deployment