Hi Taher, Currently, Flink (SQL) CDC doesn't support automatic schema change and doesn't support to consume schema change events in source. But you can upgrade schema manually, for example, if you have a table with columns [a, b, c], you can define a flink table t1 with these 3 columns. When you add new column in source RDBMS, the Flink SQL job on t1 should work fine if you are using format 'debezium-json' or 'debezium-avro-confluent', because they supports schema compatibility. When you are notified there is a schema change in the source RDBMS, then you can upgrade your Flink SQL DDL and job to include the added column, and consume from the previous savepoint [1].
Best, Jark [1]: https://ci.apache.org/projects/flink/flink-docs-release-1.13/docs/dev/table/sqlclient/#start-a-sql-job-from-a-savepoint On Wed, 5 May 2021 at 13:34, Taher Koitawala <taher...@gmail.com> wrote: > Hi All, > I have a CDC use case where I want to capture and process debezium > logs that are streamed to Kafka via Debezium. As per all the flink examples > we have to pre create the schema of the tables where I want to perform a > write. > > However my question is what if there is an alter table modify column data > type query that hits the source RDBMS, how does flink handle that schema > change and what changes are supported. If someone can give a full example > it will be very very helpful. > > > Regards, > Taher Koitawala >
