// Benutzung · Monitoring & Alerting
Muting an alert
Muting stops notifications for a time window without deleting the rule. The alert still shows in the UI while muted — you just won't be paged. The "resolved" all-clear is never muted.
Quick mute from a firing alert
On Alerts → Active, a firing alert has a Mute 1h button. It snoozes that alert's rule for 60 minutes — useful when you're already aware and working the issue.
What can be muted
Mutes match on one of three scopes:
rule — silence a specific rule everywhere.
resource — silence anything about one resource (matched on its
resource_id), across rules. Give the mute the resource'slocationandcluster(the alert's labels of those names, e.g.cluster:pve-a) to silence that one resource only: two clusters in one location can both have avm-100. When both are set, both must match. A mute without them matches the id in every location and cluster, as every mute made before rc.3 does.Scope a mute only for Proxmox and Kubernetes resources, whose ids can repeat across clusters or locations; only they carry a
clusterlabel, so a cluster-scoped mute on anything else never matches. Check ids are unique: leave check mutes unscoped. A check alert'slocationis the location id of one failing vantage, and it changes when another vantage is the one picked.POST /api/mutes { "kind": "resource", "target": "vm-100", "location": "homelab", "cluster": "cluster:pve-a", "minutes": 60 }location — silence a whole location for the window.
A mute has an expiry; once it passes, notifications resume automatically. Muting does not resolve the alert — if the condition is still breached when the mute ends, it notifies again.
Removing a mute early
Delete the mute window to resume notifications immediately (mutes are listed and removable via the alerts API; a delete restores paging for that scope).
Creating and deleting a mute is recorded in the audit log, with who did it; while it cannot be recorded the mute is refused with 503 and nothing is changed.