ccding edited a comment on pull request #10405: URL: https://github.com/apache/kafka/pull/10405#issuecomment-812744839
> It's interesting that the absolute throughput dropped significantly with 10MB segments compared with 1GB segment. I am thinking it is the cost of the extra flush. We have one extra flush per segment, which is 1 extra flush per 10,000 records for 10MB segments and 1KB records. If it were 1GB segments, there is 1 extra flush per 1,000,000 records: 1/100 of the amortized extra cost. -- 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