TODOs

• Run pipelines for all operators without submitting
• deploy operators from catalogs
• run selected tests
  • tests for secret-operator
  • tests for listener-operator
  • tests for zookeeper-operator
  • tests for opa-operator
  • tests for spark-k8s-operator
  • kerberos tests for hdfs-operator
  • gitsync test for airflow-operator
  • smoke test for kafka-operator
  • smoke test for trino-operator. 🔴 Failed. Was a problem with resources.
  • tests for druid-operator @razvan
  • tests for commons-operator @adwk67
  • tests for hive-operator @razvan
  • tests for hbase-operator @razvan
  • tests for nifi-operator @adwk67
  • tests for superset-operator @adwk67
• Run pipelines for all operators and submit and check PRs
• check certified ops that all is present and correct

Follow-up for the next release

• add support for listener and secret to the build-manifests.py script and remove the build-manifests.sh script.
  ✅ remove individual product listings from the RH portal with the exception of the general platform one

Clarify with RedHat

• What is the problem here and what is the correct solution ?
• Targeting a new OS version leads to a pipeline failure. Figure out how a new operator version can replace an old operator version in a catalog that doesn't contain the old operator. Example: commons 24.11 replaces 24.7.0 in the v4.16 catalog. But that catalog doesn't know about 24.7.0. because 24.7.0 only supports v4.11 to v4.15.
• clarify the operator-pipeline differences we see between us and what the certified operators repo is doing.
• a pipeline that has been closed can be automatically re-opened without warning: e.g. here

(these points were submitted to Redhat on 09.08.2024 at 11:45 via the certification feedback form).

Incidents

• The PR for the listener operator had a failure. It was closed by us but reopened automatically and merged,
• Now the operator is not in the catalog.