Re: Solr Performance with Ram size variation

2015-04-17 Thread Otis Gospodnetic
Hi, Because you went over 31-32 GB heap you lost the benefit of compressed pointers and even though you gave the JVM more memory the GC may have had to work harder. This is a relatively well educated guess, which you can confirm if you run tests and look at GC counts, times, JVM heap memory pool

Re: Solr Performance with Ram size variation

2015-04-17 Thread Shawn Heisey
On 4/17/2015 8:14 PM, Kamal Kishore Aggarwal wrote: Hi, As per this article, the linux machine is preferred to have 1.5 times RAM with respect to index size. So, to verify this, I tried testing the solr performance in different volumes of RAM allocation keeping other configuration (i.e Solid Sta

Re: Solr Performance with Ram size variation

2015-04-17 Thread Puneet Pawaia
Hi, This may be irrelevant but your machine configuration reminded me of some reading I had done some time back on memory vs ssd. Do a search on solr ssd and you should get some meaningful posts. Like this one https://sbdevel.wordpress.com/2013/06/06/memory-is-overrated/ Regards Puneet On 18 Apr 2

Solr Performance with Ram size variation

2015-04-17 Thread Kamal Kishore Aggarwal
Hi, As per this article, the linux machine is preferred to have 1.5 times RAM with respect to index size. So, to verify this, I tried testing the solr performance in different volumes of RAM allocation keeping other configuration (i.e Solid State Drives, 8 core processor, 64-Bit) to be same in bot