12. Day-2 operations

After your first deploy, these commands cover most ongoing work.

Manual job control

Requires a prior deploy (workers have current runner.py and update_seq).

maand job status api
maand job status api --allocations 10.0.0.1

maand job stop api --allocations 10.0.0.2
maand job start api --allocations 10.0.0.2
maand job restart api --health_check

maand job run api --target reload

These invoke runner.py → make <target> on workers over SSH.

Reference: job.md.

Ad-hoc remote commands

maand run_command "uptime"
maand run_command "df -h /opt/worker" --workers 10.0.0.1
maand run_command "docker ps" --labels web --concurrency 2

Working directory on workers: /opt/worker/<bucket_id>/.

Reference: run-command.md.

Drain for maintenance

1. Edit workspace/disabled.json
2. maand build
3. maand deploy — stops disabled allocations, does not restart them

Re-enable: remove entry, build, deploy.

Guide: disable-and-drain.md.

Remove jobs or workers

1. Delete from workspace/ (remove job dir or host from workers.json)
2. maand build — marks allocations removed
3. maand deploy — stop, remove deployed files; keeps data/ and logs/
4. maand gc — purge catalog rows and delete worker runtime trees

Guide: day-2-ops.md.

Upgrades and rollouts

# edit manifest version or job files
maand build
maand deploy --dry-run
maand deploy --jobs api

Rolling behavior: rolling-deploy.md.

Worker OS reboot: worker-reboot.md.

Debugging checklist

Guide: debugging-deploy.md.

Inspect state (quick reference)

maand info
maand cat workers
maand cat jobs
maand cat allocations
maand cat deployments
maand cat certs --jobs api
maand cat kv --jobs api

You have completed the tour