v4.30.0

Added

• Add alerts for karpenter issues.

v4.29.0

Changed

• Increase time to trigger PromtailRequestsErrors alert from 15 to 25m.

v4.28.0

Added

• Add alert to monitor the KubeadmConfig CRs having trouble generating bootstrap data.

Changed

• Ignore HelmReleases in e2e test organization namespaces for cabbage FluxHelmReleaseFailed (cilium, network-policies, coredns)

v4.27.0

Added

• KongProductionDeploymentNotSatisfied to alert on clusters starting with p.
• KongNonProdDeploymentNotSatisfied to alert on clusters not starting with p.

Removed

• Split KongDeploymentNotSatisfied into KongProductionDeploymentNotSatisfied and KongNonProdDeploymentNotSatisfied to be able to control alerting in- and outside business hours.

v4.26.2

Changed

• Remove label_replace from app_operator_app_info based alerts and use the cluster_id from the metric on CAPI.

Added

• Add cloud-provider-controller.rules to monitor the cloud-provider-controller components across providers.
• Add alerts to monitor the HelmReleases for cilium and coredns.
• Add alert to monitor the HelmRelease for the vertical-pod-autoscaler-crd app.
• Add alert to monitor Shield pods restarts.
• Add MimirRulerTooManyFailedQueries alert to detect when Mimir ruler is failing to evaluate rules

Fixed

• Fix dashboard link for MimirContinuousTestFailing alert
• Fix tests so they fail if some helm template fails to render

v4.26.1

Changed

• MimirObjectStorageLowRate and LokiObjectStorageLowRate only check management cluster apps
• MimirObjectStorageLowRate and LokiObjectStorageLowRate are less sensitive

v4.26.0

Changed

• Bump alloy-rules app version to 0.7.0
  • Upgrades alloy to 1.4.2 to 1.5.0

Added

• new MimirObjectStorageLowRate alert
• new LokiObjectStorageLowRate alert

v4.25.0

Changed

• Mimir compactor alert: better failure detection

Added

• Add new mimir continuous test alerts:
  • MimirContinuousTestFailingOnWrites
  • MimirContinuousTestFailingOnReads
  • MimirContinuousTestMissing
  • MimirContinuousTestFailing

Removed

• Remove the mimir.enabled property to replace it with the MC flavor as all CAPI MCs now run Mimir.

v4.24.1

Fixed

• Fix MonitoringAgentDown to page when both prometheus-agent and alloy-metrics jobs are missing.

v4.24.0

Added

• Add a set of sensible alerts to monitor alloy.
  • AlloySlowComponentEvaluations and AlloyUnhealthyComponents to report about alloy component state.
  • LoggingAgentDown to be alerted when the logging agent is down.
  • LogForwardingErrors to be alerted when the loki.write component is failing.
  • LogReceivingErrors to be alerted when the loki.source.api components of the gateway is failing.
  • MonitoringAgentDown to be alerted when the monitoring agent is down.
  • MonitoringAgentShardsNotSatisfied to be alerted when the monitoring agent is missing any number of desired shards.

Changed

• Update DeploymentNotSatisfiedAtlas to take into account the following components:
  • observability-operator
  • alloy-rules
  • observability-gateway
• Move all grafana-cloud related alerts to their own file.
• Move all alloy related alerts to the alloy alert file.
• Rename and move the following alerts as they are not specific to Prometheus:
  • PrometheusCriticalJobScrapingFailure => CriticalJobScrapingFailure
  • PrometheusJobScrapingFailure => JobScrapingFailure
  • PrometheusFailsToCommunicateWithRemoteStorageAPI => MetricForwardingErrors