Release notes for version 0.92.0

Release date: 07 Sep 2026

• defaultRules.groups[*].enabled may now be a template expression (e.g. {{ .Values.kubeScheduler.enabled }}), like defaultDashboards.dashboards[*].enabled already was. Disabling kubeApiServer, kubeControllerManager, kubeScheduler, or kubelet now also disables their alerting rule groups, so alerts no longer fire on scrape targets that were never enabled. See #3197
• fixed sync-job's alertmanager.rules group keeping upstream kube-prometheus's job="alertmanager-main",namespace="monitoring" selectors, which never match this chart's VMAlertmanager. The sync-job now rewrites both to a regexp matching this chart's VMAlertmanager(s) via a new labelRewrites option (alongside the existing jobNamespaces), configurable under defaultRules.labelRewrites / defaultRules.groups.<name>.labelRewrites; values may be template expressions (e.g. {{ .Release.Namespace }}), rendered before use. See #11517
• bump version of VM components to v1.151.0
• bump operator dependency chart to version 0.67.3
• fixed plugins section missing from GrafanaDatasource CRs for datasources whose type requires a Grafana plugin not already listed in grafana.plugins. See #3181
• added global.versions (metrics/logs/traces/anomaly/alertmanager), shared automatically with the embedded operator chart's own global.versions, so CRs created outside this chart default to the same version the chart deploys instead of the operator's own baked-in default. See #3186