Great suggestion, I took a look and it seems pretty useful. As a follow up question, did you do anything to disable Solr caching for certain tests?
-mike On Tue, Oct 27, 2009 at 8:14 PM, Joshua Tuberville < joshuatubervi...@eharmony.com> wrote: > Mike, > > For response times I would also look at java.net's Faban benchmarking > framework. We use it extensively for our acceptance tests and tuning > excercises. > > Joshua > > On Oct 27, 2009, at 1:59 PM, Mike Anderson wrote: > > > I've been making modifications here and there to the Solr source > > code in > > hopes to optimize for my particular setup. My goal now is to > > establish a > > descent benchmark toolset so that I can evaluate the observed > > performance > > increase before deciding rolling out. So far I've investigated > > Jmeter and > > Lucid Gaze, but each seem to have pretty steep learning curves, so I > > thought > > I'd ping the group before I sink a good chunk of time into either. > > My ideal > > performance metrics aren't so much load testing, but rather response > > time > > testing for different query types across different Solr > > configurations. > > > > If anybody has some insight into this kind of project I'd love to > > get some > > feedback. > > > > Thanks in advance, > > Mike Anderson > >
