Drain Host

Draining a host migrates all of its stacks to other hosts in a safe order, then marks the host as unschedulable. Use it before kernel updates, hardware maintenance, or retiring a host permanently.

Starting a drain

Hosts → select host → Drain
Pick one or more destination hosts (dockmesh will distribute by available resources)
Choose an ordering strategy:
- Smallest first — good for quick wins, most stacks move fast
- Largest first — frees the most resources up front
- Dependency-aware — stacks that reference each other’s volumes or networks move together

dockmesh runs the standard Migration pre-flight for each stack before starting.

During the drain

The UI shows a live progress view:

Current stack being migrated
Queued stacks with their predicted destination
Stacks that failed pre-flight (with reason)
Total bytes transferred and time remaining

Failed stacks are skipped — the drain continues with the rest. At the end you get a summary showing what succeeded, what was skipped, and why.

Mark-unschedulable-only mode

If you just want to stop new stacks from being created on a host without migrating existing ones (e.g. you’re about to reboot and want things to resettle afterwards), use Cordon instead of Drain. Cordon flips the host’s scheduling flag without touching workloads.

Uncordoning

After maintenance, Hosts → Uncordon makes the host schedulable again. Stacks you drained off earlier are not automatically migrated back — if you want them back, use Migration per stack, or keep them where they are.

Audit log

Every drain event — started, per-stack migration, completed, failed — is written to the audit log with user, timestamp, source, destination, and outcome.