Observability
Changed
- Set rollout strategy to Recreate for prometheus-operator deployment.
- Upgraded chart dependency to kube-prometheus-stack-56.21.2
- prometheus-node-exporter upgraded from 4.26 to 4.30
- thanos upgraded from 0.34.0 to 0.34.1
- alertmanager upgraded from 0.26.0 to 0.27.0
- prometheus upgraded from 2.47.1 to 2.50.1
Changed
- upgrade
prometheus-agent to 0.6.9.
Changed
- Upgraded upstream chart from 5.43.4 to 5.43.5 - see changelog for more information.
Added
- Add mimir mixins dashboards for read.
Changed
- Upgraded upstream chart from 5.43.1 to 5.43.4 - see changelog for more information.
- Remove custom CNPs and use upstream ones instead.
Added
- Add mimir mixins dashboards for write.
Changed
- Change
managed-apps-efk-stack-app dashboard owner to Atlas. - Updated
k8s-resources-node dashboard.
Changed
- use new multi-tenant-proxy, named
grafana-multi-tenant-proxy
Changed
- Added a default scrape interval value of 30s for service monitors (if they’re used).
Fixed
- Fix Prometheus - Mimir Cost Comparison dashboard.
- Moved Dashboards from deprecated ludacris team to turtles team tags.
- Removed Dashboards with deprecated data sources.
Changed
- Enable ciliumNetworkPolicies by default. If your cluster does not support cilium, disable this in the values or postpone the upgrade until cilium is installed.
Fixed
- Helm chart generation when loki is disabled and CNPs are enabled
