Overview

A collection of systemd units that attempt to fix any issues with running services. Example pipeline:

• timer
• healthcheck service
• OnFailure= attempting to fix the issue

Usability of this approach (compared to fully hand-made healthcheck orchestrator) depends on how flexible systemd is.

Workflow for adding checks is looking for simple fixes to easily observable issues after doing them by hand and adding them in.

Example checks/heals include:

• btrfs rebalance when chunk utilization is too low
• specific service restarts. but only specific issues, not secret expiration, etc - that should be handled elsewhere