Guides

Define and start instances: commands, environment, ports and sizing.

Configuration files from the host, volumes that outlive instances, and scratch space.

Check that a workload is well, use an image’s own check, and restart one that is not.

Restart instances that end on their own, and delete ones that were only ever temporary.

Freeze a running instance to disk and put it back exactly as it was.

Run commands in a guest, copy files in and out, read its console and wait for it to end.

Pull images ahead of time, use private registries, update tags and reclaim disk space.

Serve the API over TCP with TLS, and point the command line at another host.

How much CPU and memory instances may be given, overcommit and the reserve, and what is in use.

Prometheus metrics, events, and the daemon’s log and audit.

Restart and upgrade dicerd without stopping guests, back it up, and remove it.

Why an instance will not start or stay up, and how to find out.

Call the gRPC API from Go, Python or the shell.