adixitconfluent commented on PR #17870: URL: https://github.com/apache/kafka/pull/17870#issuecomment-3056223802
> Could I understand why we remove the `partitionMaxBytes` limit in share fetch? I ask because I didn't see it's mentioned anywhere in the KIP (or maybe I missed that?). Hi @showuon, in the previous flow, we could not fetch more than 1 MB for each topic partition in the share fetch request by default although the overall SFR max bytes is 50MB. In the cases of share groups, it might be a case, that not all the partitions that are present in the share fetch request can be acquired which can lead to under utilisation of request. Thus, to prevent it from happening, `partitionMaxBytes` is decided on the broker. As of now, it contains a `UNIFORM` strategy that divides overall SFR max bytes equally among the acquired topic partitions. For example - If SFR max bytes = 50MB and there are 5 acquired topic partitions, then total data that can be fetched for each topic partition from replica manager can be upto 10MB each. -- 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: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
