Details
-
Bug
-
Status: Resolved
-
Critical
-
Resolution: Fixed
-
None
Description
Currently, there are a lot of TableSink connectors uses dataStream.addSink() without setParallelism explicitly. This will use default parallelism of the environment. However, the parallelism of input transformation might not be env.parallelism, for example, global aggregation has 1 parallelism. In this case, it will lead to data reorder, and result in incorrect result.
Attachments
Issue Links
- links to