This release focuses on three major areas: infrastructure modernization, AI/ML building blocks (Kserve), and critical bug fixes. The infrastructure updates primarily address Bitnami's breaking changes by migrating to official images and Helm charts.

📘

For a full list of changes, review the release notes on GitHub: Release v4.12.0.

New

• Added support for Kubernetes v1.34. Supported Kubernetes versions are now displayed on the platform upgrade page.
• Added the following apps: Kserve (0.15.2) and Valkey (8.1.3).

Changed

• Updated core applications. See the Core application changes table below for details.

Fixed

• Resolved issue with pulling source code from a private git repository that occurred when building a container image.

Core application changes

Post-upgrade actions

After upgrading to v4.12.0, you need to perform a few additional steps to prevent the following issues.

• Cannot access the App Platform Console (web UI) when upgrading from v4.11.3 (or earlier versions). To resolve this, use kubectl to update the otomi configuration and restart several services. See Access the App Platform Console.
• CloudNative PG collation version mismatch errors: To resolve this, manually refresh the collation’s version on your databases and reindex them. This can also be done through kubectl. See Reindex CloudNative PG databases.

The sections below provide more information about these issues as well as the commands needed to resolve them.

Access the App Platform Console

After upgrading from v4.11.3 (or earlier versions) to v4.12.0, the App Platform Console (Web UI) may be unavailable. If this occurs, you can bring the console back up using kubectl. Run the following commands:

kubectl patch configmap -n otomi otomi-status --type=json -p='[{op: "replace", path: "/data/version", value: "4.11.0"}]' 
kubectl rollout restart deployment -n apl-operator apl-operator 
# Wait for ~2 minutes
kubectl rollout restart deployment -n ingress ingress-nginx-platform-controller

Reindex CloudNative PG databases

CloudNative PG workloads using the image catalog (no manually set imageName) automatically receive upgraded base images. For these workloads, a collation version mismatch warning is logged after each connection. Clients will also see a warning message. To resolve these warnings, refresh your databases collation’s version and then reindex.

• For platform databases, reindexing is done automatically.
• For all other Cloudnative PG clusters, run the following SQL commands for each database (including postgres) to refresh the collation version and reindex:
  SQL

  REINDEX DATABASE <database name>;
  ALTER DATABASE <database name> REFRESH COLLATION VERSION;
  

  You can also do this through kubectl. The commands below are a one-step cluster reindex example:
  SQL

  kubectl get cluster -n <team namespace> <cluster name> -o jsonpath='{.status.currentPrimary}'
  kubectl exec -n <namespace> <primary pod> -- psql -d postgres -S -c 'REINDEX DATABASE postgres' -c 'ALTER DATABASE postgres REFRESH COLLATION VERSION'
  kubectl exec -n <namespace> <primary pod> -- psql -d <cluster name> -S -c 'REINDEX DATABASE <cluster name>' -c 'ALTER DATABASE <cluster name> REFRESH COLLATION VERSION'