Alert on unhealthy containers #126
Labels
No labels
audit-work
bug
docs
general-admin
major-upgrade
needs-vps-sync
new-service
on-hold
outside-work
post-podman
renovate
upstream
vps
No milestone
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set
Reference
mysticalsoap/docker#126
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Monitoring.md already admits it: "a failing Docker healthcheck marks a container unhealthy but nothing scrapes or alerts on that today." The blackbox internal probes cover HTTP apps; a container that is running-but-unhealthy by its own healthcheck's definition (auth failure inside, dead worker, full disk) alerts nothing.
House-pattern fix: a docker-state exporter behind its own GET-only socket-proxy (
/containers/jsononly — same shape as promtail-socket-proxy), plus one alert onunhealthy. Explicitly not autoheal — restarting an unhealthy container masks the signal, and the goroutine-leak history (#117) argues against adding recurring exec/restart volume.