[ https://issues.apache.org/jira/browse/KAFKA-5178?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15998316#comment-15998316 ]
huxi commented on KAFKA-5178: ----------------------------- I suggest splitting the test into multiple sub-testings in order to find out the bottleneck. The roundtrip latency contains many parts of the time, such as remote time, queue time on the producer side and polling time on the consumer side. It's hard to tell which part contributes most when hitting the performance limit. Besides, I notice a fact that there are totally 64 partitions distributed on three brokers. How many physical disks does your broker machine have averagely? Could you also monitor the disk usage% during the test to see if it's too busy? > Potential Performance Degradation in Kafka Producer when using Multiple > Threads > ------------------------------------------------------------------------------- > > Key: KAFKA-5178 > URL: https://issues.apache.org/jira/browse/KAFKA-5178 > Project: Kafka > Issue Type: Bug > Reporter: Ben Stopford > > There is evidence that the Kafka Producer drops performance as we increase > the number of threads using it. > This is based on some benchmarking done in the community. I have not > independently validated these results. -- This message was sent by Atlassian JIRA (v6.3.15#6346)