VictoriaMetrics/operator v0.52.0

on GitHub

v0.52.0

Release date: 21 Jan 2025

• Dependency: vmoperator: Updated default versions for VM apps to v1.109.1 version

• Dependency: vmoperator: Updated default version for VictoriaLogs app to v1.6.1 version

• FEATURE: vmscrapeconfig: allow multiple values for zone at gceSDConfigs. See this issue for details.

• FEATURE: vmoperator: decrease latency of generated configuration updates. Previously, configuration was update after status of child objects were changed. It could take significant time at large scale. See this issue for details.

• FEATURE: vmoperator: reduce load on Kubernetes API server at prometheus-converter client.

• FEATURE: vmoperator: change default value for client.qps=50 and client.burst=100 in order to improve operator performance on scale. See this issue for details.

• FEATURE: vmoperator: add new flag controller.statusLastUpdateTimeTTL=1h to control staleness detection at status.conditions field. If operator serves large amount of object ( > 5_000) value for it should be increased.

• FEATURE: vmoperator: add more context to the failed status with the reason and logs of crashed container. See this issue for details.

• FEATURE: vmoperator: Properly expose service ports for VMServiceScrape generated for CRD objects by operator. Add -vmbackupmanager suffix to the VMServiceScrape job name and expose only well-known http port. See this issue for details.

• BUGFIX: vmagent: properly build relabelConfigs with empty string values for separator and replacement fields. See this issue for details.

• BUGFIX: converter: properly format regex single value expression at Prometheus Operator CRD relabelings and metricsRelabelings. See this issue for details.

• BUGFIX: vmoperator: keep spec.template.annotations added by 3rd party tools, kike kubectl rollout restart. See this issue for details.