We did a bunch of sampling to no particular aid, broadly speaking the answer was "it's doing a bunch of talking".
For those who might want to know what this was in the end, during part of our previous debugging we enabled `javax.net.debug=all` and didn't twig that that had no effect on the log4j logs, and didn't notice the vast number of iops to `kafkaServer.out`. Writing that log was eating all the CPU. On 23 September 2017 at 00:44, jrpi...@gmail.com <jrpi...@gmail.com> wrote: > One thing worth trying is hooking up to 1 or more of the brokers via JMX > and examining the running threads; If that doesn't elucidate the cause, > you could move onto sampling or profiling via JMX to see what's taking up > all that CPU. > > - Jordan Pilat > > On 2017-09-21 07:58, Elliot Crosby-McCullough <elliot.crosby-mccullough@ > freeagent.com> wrote: > > Hello, > > > > We've been trying to debug an issue with our kafka cluster for several > days > > now and we're close to out of options. > > > > We have 3 kafka brokers associated with 3 zookeeper nodes and 3 registry > > nodes, plus a few streams clients and a ruby producer. > > > > Two of the three brokers are pinning a core and have been for days, no > > amount of restarting, debugging, or clearing out of data seems to help. > > > > We've got the logs at DEBUG level which shows a constant flow much like > > this: https://gist.github.com/elliotcm/e66a1ca838558664bab0c91549acb251 > > > > As best as we can tell the brokers are up to date on replication and the > > leaders are well-balanced. The cluster is receiving no traffic; no > > messages are being sent in and the consumers/streams are shut down. > > > > From our profiling of the JVM it looks like the CPU is mostly working in > > replication threads and SSL traffic (it's a secured cluster) but that > > shouldn't be treated as gospel. > > > > Any advice would be greatly appreciated. > > > > All the best, > > Elliot > > >