Operations commands
Observability
Section titled “Observability”hyperlake observability sampleshyperlake observability query --cluster CLUSTER_ID --source loki --sql SQLhyperlake observability query --cluster CLUSTER_ID --source prometheus --sql SQLThe source is a governed observability catalog. Loki presets can target workload logs, OPA decisions, lineage, or Trino logs. Prometheus queries can select metrics, labels, windows, and limits.
Governed services
Section titled “Governed services”hyperlake service listhyperlake service describe --service SERVICE_IDhyperlake service execute --cluster CLUSTER_ID --service SERVICE_ID --action ACTION_ID --parameters '{}'hyperlake service status --run RUN_IDhyperlake service approve --run RUN_IDhyperlake service cancel --run RUN_IDhyperlake service rollback --run RUN_IDOnly server-allowlisted services and actions are executable. An operation may pause for approval. Use the canonical service ID returned by service list, not a display label or installation UUID.
Monitors and diagnostics
Section titled “Monitors and diagnostics”hyperlake monitor listhyperlake monitor run --id MONITOR_IDhyperlake monitor history --id MONITOR_IDhyperlake k8sgpt analyze --cluster CLUSTER_ID --filter Pod --explainMonitor definitions can include expectations, schedules, and notifications. K8sGPT runs through the managed app/cluster identity path; it does not require users to paste kubeconfig or cloud private keys into the CLI.