-
Notifications
You must be signed in to change notification settings - Fork 3.3k
Open
Labels
Description
Description
Three deprecated Kafka topics are still defined in the codebase but are no longer actively used:
MetadataAuditEvent_v4MetadataGraphEvent_v4MetadataSearchEvent_v4
These topics were deprecated mid-2022 when DataHub transitioned to the more flexible MetadataChangeLog (MCL) and MetadataChangeProposal (MCP) event model. The legacy code remains only for backward compatibility, but it creates confusion for users and adds maintenance overhead.
Current State
Code References
- Topic constants defined in
Topics.java - Topic configuration in
TopicConventionFactory.java - Explicitly disabled in
application.yaml:metadataAuditEvent.enabled: false - Schema ordinals maintained in
EventSchemaConstants.javaandSchemaIdOrdinal.javafor deserialization only
No Active Usage
MetadataAuditEvent_v4: TheproduceMetadataAuditEventmethod exists but has zero call sites in the codebaseMetadataGraphEvent_v4andMetadataSearchEvent_v4: Originally designed as CDC streams but never actively produced or consumed- All three topics are marked as deprecated in documentation
Related Issues/PRs
- Original MAE deprecation (mid-2022)
- GitHub Issue What are
MetadataSearchEventandMetadataGraphEventfor? #1696 (referenced in code comments) - GitHub PR docs(kafka): clarify Kafka topic retention requirements #15149 (MetadataGraphEvent_v4 schema removal)
Additional Context
Recent user question from Slack that prompted this issue:
"Why are MetadataGraphEvent_v4 and MetadataSearchEvent_v4 still preserved in the source code if they are not used anymore? Do I need that topic to be available while deploying datahub?"
This confusion will persist until we fully remove the deprecated code.