[jira] [Created] (FLINK-34424) BoundedBlockingSubpartitionWriteReadTest#testRead10ConsumersConcurrent times out

Matthias Pohl created FLINK-34424: - Summary: BoundedBlockingSubpartitionWriteReadTest#testRead10ConsumersConcurrent times out Key: FLINK-34424 URL: https://issues.apache.org/jira/browse/FLINK-34424 Pr

Making CollectSinkFunction to wait till all results are consumed

Hey all, We propose to slightly change the behavior of the CollectSinkFunction and make it wait till all

[jira] [Created] (FLINK-34423) Make tool/ci/compile_ci.sh not necessarily rely on clean phase

Matthias Pohl created FLINK-34423: - Summary: Make tool/ci/compile_ci.sh not necessarily rely on clean phase Key: FLINK-34423 URL: https://issues.apache.org/jira/browse/FLINK-34423 Project: Flink

Re: [DISCUSS] Alternative way of posting FLIPs

+1 for option 1. Thanks Martijn for taking care of this, I've also filed a issue[1] for the infra a few days ago, but haven't been resolved yet. [1] https://issues.apache.org/jira/browse/INFRA-25451 Jane Chan 于2024年2月10日周六 14:33写道： > > +1 for option 1. > > Hi Yun, > > > I am concerned about whet

Re: [DISCUSS] FLIP 411: Chaining-agnostic Operator ID generation for improved state compatibility on parallelism change

FWIW, there is a mechanism in table/SQL[1] that can generate uids based on the type of physical node, if we set it to "ALWAYS", then it's not limited to compiled plan (maybe edited by users). I haven't tried this in production yet, but I much like the idea of keeping the uids of SQL jobs stable, e