1. Docs
  2. Changes and releases
  3. Observability

Observability

  • Observability mimir-app v0.30.0

    Changed

    • Upgrade mimir-distributed chart to version 6.1.0.

    Fixed

    • Remove the unused enterprise/legacyLabels branches from the mimir.labels, mimir.podLabels and mimir.selectorLabels helpers, aligning them with upstream. The mimir-distributed 6.1.0 chart dropped the default enterprise block from its values, which made these helpers fail with a nil-pointer error during templating.

    Removed

    • Remove the prometheus.io/service-monitor: "false" label from the gateway service, reverting the 0.29.1 bugfix. It is no longer needed as the mimir-distributed 6.1.0 chart removes the gateway serviceMonitor upstream.
  • Observability tempo-app v0.18.0

    Changed

    • Fix ABS config to not override AppVersion in Chart.yaml
    • Upgrade Tempo Vulture chart from 0.12.10 to 0.13.0
      • Bumps the tempo-vulture image from 2.10.5 to 3.0.2, which is the only change in the chart
      • Pins -tempo-metrics-backoff-duration=0s: the TraceQL metrics check went from disabled to enabled (every 10s) upstream, and it cannot pass here because we do not enable the metrics-generator local-blocks processor
      • Drops support for the -rhythm-rf1-after flag (unused here) and starts honouring -tempo-push-tls, which was previously ignored (we leave it disabled)
    • Upgrade Tempo chart from 2.23.0 to 2.26.2
      • Upgrades Tempo from 2.10.5 to 2.10.7
      • Bumps the memcached image from 1.6.42-alpine to 1.6.43-alpine
      • Bumps the rollout-operator sub-dependency from 0.49.0 to 0.50.0 (not enabled here)
      • The metrics-generator now gets a livenessProbe (/ready, 60s initial delay), which it did not have when running as a StatefulSet (chart 2.25.4)
      • Ingester and metrics-generator moved onto the shared pod template; container ports are reordered and the ingester grpc port is now declared via ingester.extraPorts. Rendered output is otherwise unchanged, so this is a rolling restart only
      • Adds (disabled by default) ingest/Kafka write path with the experimental block-builder and live-store components, plus ingester.enabled to turn the ingester off (chart 2.24.0)
      • Adds a dnsConfig value on every component with defaults/tempo fallback, deprecating dnsConfigOverides (chart 2.26.1)
      • Adds distributor.shutdownDelay for graceful load-balancer draining, disabled by default (chart 2.26.0)
  • Observability mimir-app v0.29.1

    Fixed

    • disabled gateway metrics servicemonitor. Had been added in 0.29.0 but not working.
  • Observability mimir-app v0.29.0

    Changed

    • Fix ABS config to not override AppVersion in Chart.yaml
    • Upgrade chart to version 6.0.6
    • Upgrade mimir to 3.1.4
  • Observability observability-bundle v3.3.0

    Added

    • Add KSM metrics for Gateway API ListenerSet and ReferenceGrant resources.

    Changed

    • Update Gateway API KSM configs to v1 for Gateway, GatewayClass, HTTPRoute, GRPCRoute, TLSRoute and BackendTLSPolicy.
  • Observability dashboards v4.28.0

    Changed

    • Update Kafka dashboards to reload variable on time range changes and use shared crosshair.
  • Observability dashboards v4.27.0

    Added

    • Add new Audit Volume dashboard to observe size of audit logs by cluster.
  • Observability observability-bundle v3.2.0

    Changed

    • Update kube-prometheus-stack and prometheus-operator-crd to 22.0.0
    • Update alloy-app to 0.21.0
  • Observability kube-prometheus-stack-app v22.0.0

    Changed

    • Upgraded chart dependency to kube-prometheus-stack-87.3.0 (from 85.2.1)
      • prometheus-operator to 0.92.0 (via 0.91.0)
      • alertmanager image tag from v0.32.1 to v0.33.0
      • prometheus image tag from v3.11.3-distroless to v3.12.0-distroless
      • grafana subchart from 12.3.3 to 12.7.1
      • kube-state-metrics subchart from 7.4.0 to 7.5.1
      • kube-webhook-certgen image tag from 1.8.3 to 1.8.4
      • Refreshed bundled default Prometheus alerting/recording rules and Grafana dashboards.
      • prometheus.prometheusSpec.containers and initContainers are now rendered through tpl (87.3.0), so templated values inside them are evaluated. We do not set these values, so no impact.

    Note: Update the matching-major prometheus-operator-crd app (to the v0.92.0 CRDs) before/alongside this upgrade. The PrometheusTopologySharding and PrometheusShardRetentionPolicy feature gates are now enabled by default in v0.92.0.

  • Observability prometheus-operator-app v22.0.0

    Changed

    • Upgraded chart dependency to kube-prometheus-stack-87.3.0 (from 85.2.1)
      • prometheus-operator to 0.92.0 (via 0.91.0)
      • alertmanager image tag from v0.32.1 to v0.33.0
      • prometheus image tag from v3.11.3-distroless to v3.12.0-distroless
      • grafana subchart from 12.3.3 to 12.7.1
      • kube-state-metrics subchart from 7.4.0 to 7.5.1
      • kube-webhook-certgen image tag from 1.8.3 to 1.8.4
      • Refreshed bundled default Prometheus alerting/recording rules and Grafana dashboards.
      • prometheus.prometheusSpec.containers and initContainers are now rendered through tpl (87.3.0), so templated values inside them are evaluated. We do not set these values, so no impact.

    Note: Update the matching-major prometheus-operator-crd app (to the v0.92.0 CRDs) before/alongside this upgrade. The PrometheusTopologySharding and PrometheusShardRetentionPolicy feature gates are now enabled by default in v0.92.0.