Hi, I d like to do a stress testing of my solr application to see how many concurrent requests it can handle and how long it takes. But I m not sure if I have done it in proper way (likely not)... responses seem to be very slow
My configuration: 1 Solr instance, using the default settings distributed in the example code 1.1 million documents indexed duo core 1.8 GB machine with ram 2GB My testing approach: I created 100 Threads, each of which producings exactly same one query to Solr, and the query should return 7 results. The wait interval bewteen each thread is 1 second. In other words, every 1 second there is a thread created issuing same query to solr server. The time taken to serve 1 request is 2 seconds. I was expecting that response to first thread take 2 seconds, then much shorter for the rest, however, it seems results are very random. The last thread took even 10 seconds to complete. In another test I created the 100 threads instantaneously and each issuing the same query instantaneously, and for a while Solr went silent, until about 15-20 seconds it started to send responses... I suspect if this is the proper way to do the stress testing? if not may I have some pointers please? If the test design makes sense, clearly there must be something I can do to improve performance, may I have some pointers too please? Many thanks! -- View this message in context: http://www.nabble.com/Solr-search-stress-testing-tp15299687p15299687.html Sent from the Solr - User mailing list archive at Nabble.com.
