Removing the credentials would otherwise break these playbooks mid-deploy: they template uptime_kuma_password with no assert to stop them first. 100 tasks are now guarded by uptime_kuma_enabled (false), so deployments run normally and the monitoring sections skip. A further 28 tasks were already self-guarding on monitor_setup/push_url being defined; verified that a skipped task's registered variable makes those skip cleanly rather than error. The blocks are kept on purpose — the health-check logic is the durable part and should be rewired to whatever replaces Uptime Kuma. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|---|---|---|
| .. | ||
| bitcoin_knots_vars.yml | ||
| deploy_bitcoin_knots_playbook.yml | ||