3.1.0

Bug fixes:

• Crash events are now reported for apps started in different namespaces.
• Eirini workloads are now scheduled much more efficiently on Kubernetes nodes (see this story for details).
  • Deploying this release will result in a migration Job being run that will fix CPU requests for all existing workloads. If you're not deploying Eirini as part of cf-for-k8s, or generally using kapp, you should delete this Job after it has run, as it will might result in "Job already existing" errors at your next upgrade.

Other

• We have documented all our ConfigMaps.
• We have added documentation about deploying Eirini with our new plain YAML releases.
• instance-index-env-injector doesn't depend on Eirinix anymore.

CRDs (experimental)

• eirini-controller now emits Prometheus metrics.
• eirini-controller and opi now use separate ConfigMaps.

⚠️ Deprecation notice

• Given the given deprecation of Eirini in KubeCF, we will remove any KubeCF-specific components from this release with the next major release. These will include:
  • metrics-collector
  • route-collector
  • route-pod-informer
  • route-statefulset-informer