akalash commented on pull request #32: URL: https://github.com/apache/flink-benchmarks/pull/32#issuecomment-930349973
I compared this benchmark with UnalignedCheckpointTimeBenchmark and they are perfectly the same. Just in case, it is not possible to compare it directly because this benchmark shows pure checkpoint time(because we triggered it manually) while the old one has an overhead like the checkpoint interval and so on. But if we take the average pure time from the old benchmark it would be the same as in the new one. According to the buffer debloat, I don't see the big problem here, I mean it is indeed volatile but it is exactly how it works now especially with such a small number in configuration. I tried many other configurations but the result is the same. So I propose to keep this benchmark as is and then check the trend after a while. -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
