docs(agents): record the compose-only observability gap found by the reboot
Rebooting Dankclaw for the amd_iommu=off change exposed five services that exist only as compose.yaml services with no Quadlet unit (prometheus, grafana, loki, alloy, node-exporter): they did not come back and had to be restored by hand. The boot-persistence rule already requires units for them; this records the concrete gap so the next reboot is not a surprise.
This commit is contained in:
@@ -200,6 +200,16 @@ Pi-hole unit existed but crash-looped on a sysctl conflict (a later-sorted
|
||||
drop-in reset `net.ipv4.ip_unprivileged_port_start` to 80), taking LAN DNS
|
||||
down with it.
|
||||
|
||||
**Known gap (2026-09-11, found while rebooting Dankclaw for the `amd_iommu=off`
|
||||
change):** five Dankclaw services still run only from `dankclaw/compose.yaml`
|
||||
with no Quadlet unit — `prometheus`, `grafana`, `loki`, `alloy` and
|
||||
`node-exporter` (all under `~/.config/containers/systemd/` absent, but present
|
||||
in the compose file). They did not come back after that reboot and were
|
||||
restored with `./scripts/deploy.sh dankclaw --service <name>`; they will not
|
||||
survive the next one either. Giving them units is tracked work, not a solved
|
||||
problem — treat every reboot as expected to take the observability stack down
|
||||
until it is done.
|
||||
|
||||
## Testing
|
||||
|
||||
Tests are mandatory before commit. `bats` must be installed on the
|
||||
|
||||
Reference in New Issue
Block a user