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

Posted to user@storm.apache.org by Michael Schonfeld <mi...@modernmast.com> on 2015/01/06 21:45:08 UTC

DRPC Performance Issues

Hey guys,

We've recently noticed our DRPC gets extremely slow (>30s to finish
requests that normally take under 1s) when other topologies experience very
high-loads. Ideally, we'd like Storm's DRPC to power our backend API
server. Couple questions:

1. Should I even be using Storm's DRPC as our main API server?

2. If so, should I be upgrading to bigger machines? Or, should I be using
the isolation scheduler and isolate my DRPC topology?

I'd love to hear your thoughts and/or experience with this!

Thanks,
- Michael