You have a cluster — what next

At this point you have a master and at least one worker, both showing healthy in the panel. That’s a working cluster. Everything past here is optional, and you can come back to any of it later.

Run something on it

• Prepare a disk and create a storage — turn a plugged-in disk into a pool, then carve out a sized storage for an app to use.
• Run an app on a storage — install PostgreSQL, or any package, with its data on a storage you created, and remove it cleanly when you’re done with it.
• Get your connection ready and publish a service — make an application reachable from outside your network, with real, working security.

Keep it running

• Maintain a node — the day-to-day health actions available for any machine in your cluster.
• Remove a node — take a machine back out of the cluster cleanly.
• Manage and remove what’s published — check on everything you’ve published, and take it down cleanly when you’re done.
• Migrate the operating system to an SSD — move a machine’s operating system off its memory card and onto a faster, more durable disk.
• Upgrade Kubernetes — move the cluster onto a newer version of the software underneath it.
• Keep the cluster’s software up to date — how the panel and its connecting software stay current.

If you’re curious

• What Headlamp is — an optional, more detailed view of your cluster, in case you ever go looking for it or someone points you at it.

If something goes wrong

Troubleshooting is organised by symptom — what you’re seeing, not what caused it — so you can search for what’s actually happening on your screen rather than a cause you’d have to already know.