I remembered there is another implementation using lucene index file as the
look up table not the in memory FST
FST has its advantage in speed but if you writes documents during runtime,
reconstructing FST may cause performance issue
On Tue, Jan 17, 2012 at 11:08 AM, Robert Muir wrote:
> looks l
you may disable FST look up and use lucene index as the suggest method
FST look up loads all documents into the memory, you can use the lucene
spell checker instead
On Tue, Jan 17, 2012 at 10:31 AM, Dave wrote:
> I've tried up to -Xmx5g
>
> On Mon, Jan 16, 2012 at 9:15 PM,
What is the largest -Xmx value you have tried?
Your index size seems not very big
Try -Xmx2048m , it should work
On Tue, Jan 17, 2012 at 9:31 AM, Dave wrote:
> I'm trying to figure out what my memory needs are for a rather large
> dataset. I'm trying to build an auto-complete system for every
>
ab is not the best testing tool to test website performance
try loadrunner or apache httpclient instead, they act more like an browser
On Sat, Dec 19, 2009 at 11:20 AM, Olala wrote:
>
> I used ab(apache bench) to test with handle 1000 requests, with a maximum
> of
> 300 requests running concurr
