[GitHub] [kafka] chia7712 commented on pull request #10269: KAFKA-12410 KafkaAPis ought to group fetch data before generating fet…
chia7712 commented on pull request #10269: URL: https://github.com/apache/kafka/pull/10269#issuecomment-801197578 > I'm rewriting #9944 to use the autogenerated structures based on this PR. Just pushed a version that simplifies the unresolved topic ID handling. I tried to make it easier to build the fetch response using the data object. Going to try to build the response using the data object in most places today and I can push that version as soon as I can. sounds good. If the new approach is very different from #9944, please open a new PR in order to compare them :) 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
[GitHub] [kafka] chia7712 commented on pull request #10269: KAFKA-12410 KafkaAPis ought to group fetch data before generating fet…
chia7712 commented on pull request #10269: URL: https://github.com/apache/kafka/pull/10269#issuecomment-800745691 > is there any work in progress for a KafkaApis.handleFetchRequest test? I suspect it would be similar but maybe a bit harder than what I did for the LeaderAndIsr version #10071 (trading replicamanager for fetchmanager, etc). This benchmark would be helpful for #9944 as you could probably guess :) this PR is blocked by #9944. This PR (and other related issues) aim to remove all extra collection creation by using auto-generated data. In #9944 we have to create a lot of collections to handle the topic id in fetch request. Hence, I need to rethink the value (and approach) of this PR :) 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
[GitHub] [kafka] chia7712 commented on pull request #10269: KAFKA-12410 KafkaAPis ought to group fetch data before generating fet…
chia7712 commented on pull request #10269: URL: https://github.com/apache/kafka/pull/10269#issuecomment-795066426 @ijuma the JMH result of fetch session is attached. I tried to have a JMH for stress fetch. However, `KafkaApis.handleFetchRequest` is hard to be a JMH. It requires a lot of changes ... 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
[GitHub] [kafka] chia7712 commented on pull request #10269: KAFKA-12410 KafkaAPis ought to group fetch data before generating fet…
chia7712 commented on pull request #10269: URL: https://github.com/apache/kafka/pull/10269#issuecomment-794237954 > I was thinking about the jmh microbenchmarks that stress fetch, fetch session and so on. will copy that. 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
[GitHub] [kafka] chia7712 commented on pull request #10269: KAFKA-12410 KafkaAPis ought to group fetch data before generating fet…
chia7712 commented on pull request #10269: URL: https://github.com/apache/kafka/pull/10269#issuecomment-794221218 @ijuma The results of performance tests are attached. It does not show obvious performance regression. Will run more tests tomorrow. 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
[GitHub] [kafka] chia7712 commented on pull request #10269: KAFKA-12410 KafkaAPis ought to group fetch data before generating fet…
chia7712 commented on pull request #10269: URL: https://github.com/apache/kafka/pull/10269#issuecomment-792943140 > Thanks for the PR. Can you check the perf impact of these changes? sure. will add benchmark results tomorrow. 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