No, that's not true. As long as you're not limited by some resource, avgRequestsPerSecond can grow without impacting avgTimePerRequest much.
avgTimePerRequest is the elapsed time from the beginning of Solr handling the request to the end, measured in clock time. Say it takes 100 ms. Of that 100 ms, let's say (and these are examples, not real data) that 50ms is sitting around waiting for disk. A second query could be being serviced in that same interval, so at 1 QPS you'd have an averageTimePreRequest of 100ms. Ditto at 2 QPS. That said, at some point you *will* get a relationship between the two numbers, when some resource is being used 100%. At that point as your QPS rises, so will your avgTimePerRequest. Of course it's not that clean, but that's the idea... Best Erick On Thu, Nov 18, 2010 at 9:02 AM, mesenthil < senthilkumar.arumu...@mtvncontractor.com> wrote: > > We have recently upgraded some of our solr instances to 1.4.1 from 1.3. > Interestingly both these parameter values got increased after our upgrade. > When avgRequestsPerSecond increases, avgTimePerRequest should be increased. > But it is not in our case.. > > Any thoughts ? > -- > View this message in context: > http://lucene.472066.n3.nabble.com/Meaning-of-avgTimePerRequest-avgRequestsPerSecond-in-SOLR-stats-page-tp1922692p1924031.html > Sent from the Solr - User mailing list archive at Nabble.com. >
