You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@storm.apache.org by Re'em Bensimhon <re...@forter.com> on 2015/10/04 12:25:49 UTC

TCP connection threads taking a lot of CPU

Hey everybody

While analyzing a different issue on our production topology I noticed the
following in jvmtop command output:

2447 RMI TCP Connection(7)-10.203.1       RUNNABLE 37.24%     0.08%

37% CPU for TCP connections? Why is that?

We're using storm 0.9.5. Out topology consists of ~200-300 bolt instances
per node, I'm guessing that'll be related just unsure exactly why.

Anyone else seen this?