bardock commented on issue #2013: Unable to consume messages from a partition URL: https://github.com/apache/incubator-pulsar/issues/2013#issuecomment-406349051 @sijie We don't have bookkeeper's metrics. This cluster has 9 bookies, and we see that 4 of them were using more CPU than the rest (3 of them were stuck at 13%). When we restarted the cluster, every bookie's CPU went normal. ![screen shot 2018-07-19 at 12 43 10 pm](https://user-images.githubusercontent.com/1980715/42953792-910c4628-8b51-11e8-9bbe-0fc10969c757.png) The problem with partition-4 of this topic started at 16:05 (GMT -3). So we believe these bookies were performing poorly and caused the throttling. Sadly, we don't have a jstack for that moment. Is there anything we can tune up or monitor to prevent this issue? Here is our [config file](https://github.com/apache/incubator-pulsar/files/2210966/bookkeeper.conf.txt). We are using bk 4.7.0 (shipped with pulsar 2.0.0-rc1) but we are planning to upgrade to 4.7.1. Thanks!
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on 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 With regards, Apache Git Services