Yes, definitely. Our typical setup is 16Gb physical RAM and -Xmx4G per node (index size is about 1-1.5Gb per node). So there is plenty of room for OS cache, I guess. I'll take a closer look at the number of major page faults, but at the moment iostat says that everything is pretty fine.
On the other hand, it seems like we could pack those search nodes more densely on RAM (we have 20 of those), but it's the topic for another story. On Aug 8, 2013, at 11:18 PM, Michael McCandless <luc...@mikemccandless.com> wrote: > Note that you should still run a tight ship, ie don't give excess heap > to Lucene, and instead let the OS take up the slack of any spare RAM > for IO caching. --- Denis Bazhenov <dot...@gmail.com>
