You are viewing a plain text version of this content. The canonical link for it is here.

Posted to user@storm.apache.org by Eric Ruel <er...@wantedanalytics.com> on 2014/05/11 15:30:25 UTC

RE: Process Latency very higher than Execute Latency

We have a topology that use Trident..


it may be normal, but I don't understand why the Process Latency is so higher than Execute Latency (more than 10 times)

it seems that the time is spent with the stream $coord-bg0

in this example, we have an ExecuteLatency of 0.121ms vs 10867.212 ms for Process Latency for this stream


we know that there is a cost to use Trident but that much?


Is there something to do to reduce this latency or it's normal?


see screenshot.


http://postimg.org/image/qn9llvhs5/