Each task is a short guide for one job. They assume a tenant cluster that is already deployed, as in the tutorial.
Verify a deployment
Check that a tenant is running, that its network policies are in place and that it passes the CIS benchmark.
Reach the dashboards
Open Grafana, Prometheus, Alertmanager, OpenCost, Hubble and Longhorn from the controller.
Turn on an event bus
Deploy NATS, Valkey or Kafka on a tenant that needs one.
Restart a workload when its config changes
Use Reloader to roll a Deployment, StatefulSet or DaemonSet when a ConfigMap it reads is updated.
Allow a new image publisher
Add a container image source to the registry allowlist and pin the image by digest.
Restore a database to a point in time
Recover the primary's data as it was at a given moment, into a new cluster, from the backups in Garage.
Find a dropped connection
Use Hubble to see which network policy blocked a flow and from where.
Read the audit log
Find who did what at the Kubernetes API, using the audit log in Loki.
Check Falco events
Read the runtime security events Falco has raised, in Loki or straight from the pods.
Rotate a secret
Change a password or token that a platform component reads, and get the component to use it.
Upgrade a component
Move a platform component to a new version, keeping its image pinned by digest.
Rotate the proof key
Replace the key postgres signs proofs of life with, without the failover controller losing a round.
Rotate the exchange certificate
Give the failover observers a new mutual TLS certificate before the old one expires.
Add a live replica
Put a new live replica behind the current primary with one command, for example after a failover left none.
Tear down a tenant
Remove RKE2, etcd and every volume from the nodes and return them to a clean state.