Files
cozystack/packages/system/monitoring/tests
Yan Bondarenko 89028b455e fix(monitoring): route InfoInhibitor to the blackhole receiver
InfoInhibitor exists to be the source side of an Alertmanager inhibit
rule and, as its own description says, to go to a null receiver. The
platform Alertmanager config has no inhibit rule, and the rule has
carried severity major since it was imported alongside Alerta, so
Alerta shows it one level below critical. Alerta's Telegram and Slack
plugins forward it by default, and a notify_email label inherited from
the info alert gets it emailed too.

It fires whenever a rule labelled severity="info" is active with no
warning or critical alert in the same namespace and cluster. The
platform's own rules no longer use info, but rules it does not write
still do. CDI generates CDIStorageProfilesIncomplete, KubeVirt
generates KubeVirtVMGuestMemoryAvailableLow, and operators add their
own. Each of them turns into a major InfoInhibitor in Alerta.

none is the severity upstream kube-prometheus gives this rule, and the
platform config already sends severity="none" to the blackhole ahead
of every route that delivers. The severity contract refuses none
because such an alert is never delivered, so InfoInhibitor is named as
the one alert for which that is the point. Tests hold both ends, so
neither the label nor the route order can drift back to delivering it.

Assisted-by: LLM
Signed-off-by: Yan Bondarenko <202671653+yankawai@users.noreply.github.com>
2026-10-18 23:16:45 +03:00
..