Hi all, I would like to ask for a review of FLINK-40508: Support DB2 pipeline connector.
JIRA: https://issues.apache.org/jira/browse/FLINK-40508 PR: https://github.com/apache/flink-cdc/pull/4533 The PR adds a new flink-cdc-pipeline-connector-db2 module so that DB2 can be used as a source with the YAML Pipeline API. Today DB2 is only reachable from the DataStream and SQL APIs through flink-connector-db2-cdc. The new connector reuses that source connector as-is on top of the incremental snapshot framework, so nothing in flink-connector-db2-cdc changes, and it follows the same layout as the existing Postgres and Oracle pipeline connectors: a DataSourceFactory plus an EventSourceProvider, with a MetadataAccessor and a record emitter that translate Debezium events into CDC events. It covers the snapshot and incremental phases, chunk splitting with a configurable chunk key column, schema evolution, and DB2 specific types such as DECFLOAT, XML and CLOB/BLOB. The options are the ones the source connector already supports and are documented in the new docs page, which comes in both English and Chinese. Verification: - Db2TypeUtilsTest covers every supported type mapping and the unsupported type path. - Db2DataSourceFactoryITCase and Db2PipelineITCase run against a real DB2 instance through testcontainers, reusing Db2TestBase from the source connector's test-jar. - The module is registered in .github/workflows/modules.py, so CI builds and tests it on both Flink 1.20.3 and Flink 2.2.0. Those jobs, including the ITCases, are green on my fork: https://github.com/suntectec/flink-cdc/actions/runs/34694573035 Two practical notes. The workflow run on the upstream PR is still waiting for a committer to approve it, since this is my first contribution here. And FLINK-40508 is still unassigned while I do not have the permission to assign it to myself — please let me know if it should be assigned before the review moves forward. Thanks, Jagger
