zjuwangg commented on pull request #15283: URL: https://github.com/apache/flink/pull/15283#issuecomment-802786244
> And what are those certain situations? If the parallelism is a fixed value (because it runs on CI with a supposedly constant number of CPU cores) and the number of required buffers is also a fixed value (because the job didn't change), then why is it failing _sometimes_? It's very good question. I don't known too much about runtime resource allocate behavior. Is there more experience or advice to dig into deeper? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
