Common Operations

Common Operations

Day-to-day commands for managing the media stack. Sections below cover deployment (single service, all services, full site), backup, restore, migration from existing LXC containers, lint, image pruning, log inspection, and dry-run/check-mode workflows.

Deploy a single service

ansible-playbook playbooks/deploy-service.yml -e service_name=radarr

Deploy all services

ansible-playbook playbooks/deploy-services.yml

Full deployment

Provision the VM, configure the base system, and deploy all services:

ansible-playbook playbooks/site.yml

# Or step by step:
ansible-playbook playbooks/provision-vm.yml
ansible-playbook playbooks/setup-base.yml
ansible-playbook playbooks/deploy-services.yml

Backup

# Run backups for all services
ansible-playbook playbooks/backup.yml

See Backup & Restore for details on automated timers, retention, encryption, and restore procedures.

Restore

ansible-playbook playbooks/restore.yml \
  -e service_name=radarr \
  -e backup_file=/data/backups/config/radarr/radarr-2025-01-15.tar.zst.age

For encrypted backups, pass the identity file:

ansible-playbook playbooks/restore.yml \
  -e service_name=radarr \
  -e backup_file=/data/backups/config/radarr/radarr-2025-01-15.tar.zst.age \
  -e backup_age_identity_file=/path/to/age-identity.txt

Multi-container services (e.g., Open Notebook) work the same way:

ansible-playbook playbooks/restore.yml \
  -e service_name=open-notebook \
  -e backup_file=/data/backups/config/open-notebook/open-notebook-2025-01-15.tar.zst.age \
  -e backup_age_identity_file=/path/to/age-identity.txt

See Backup & Restore for full restore instructions.

Migrate from existing LXC containers

ansible-playbook playbooks/migrate.yml -e source_host=old-lxc-host

This will:

1. Create a Proxmox snapshot for rollback
2. Verify SQLite integrity on source
3. Rsync config directories to the new VM (Immich skips regenerable content: thumbnails, transcoded video, profile images)
4. Fix ownership and start services
5. Verify health checks pass

Lint

ansible-lint playbooks/ roles/
yamllint .

Image pruning

Dangling container images are cleaned up automatically by a weekly timer and after each successful autodeploy. To run a manual prune:

# Prune dangling images only
podman image prune -f

# Check disk usage
podman system df

Log inspection

# Run an inspection immediately
systemctl --user start mms-log-inspect.service

# Check the scheduled timer
systemctl --user status mms-log-inspect.timer
systemctl --user list-timers mms-log-inspect.timer

# Read the latest report
python3 -m json.tool ~/config/logging/inspection/latest-report.json

# Validate policies locally before deployment
scripts/generate-test-corpus \
  --scenario faulty \
  --entries 40 \
  --output /tmp/mms-faulty.jsonl
scripts/mms-log-inspect \
  --input-jsonl /tmp/mms-faulty.jsonl \
  --policy examples/log-policies \
  --output-json /tmp/mms-log-report.json \
  --fail-on critical

See Observability for the policy schema, notification setup, and troubleshooting notes.

Dry run

ansible-playbook playbooks/site.yml --check --diff