From what I can tell there's currently no way to control the
parallelism of the LookupTableSource. It /seems/ to be bound to the
parallelism of the input stream (i.e., the CDC stream).
I'd suggest to create a jira ticket for this.
On 11/01/2022 06:43, Jonathan Weaver wrote:
I'm attempting to d
I'm attempting to do a proof of concept conversion of a DataStream based
Flink program over to using almost entirely Table SQL.
I have a primary CDC stream (an unbounded scan table source) that does two
joins to LookupTableSource tables and then on to a sink.
In the datastream program the only wa
