On Friday 14 December 2007 16:56:07 Sean Timm wrote: > See LUCENE-997 > https://issues.apache.org/jira/browse/LUCENE-997
I actually have had this problem (ran into a synchronized{} bottleneck in FSDirectory) and "fixed" it by introducing a pool of Searchers (each allocates memory and each holds QueryFilter redundantly cached in a WeakHashMap...) and do deny queries if the pool is exhausted. The FAQ does not recommend this approach. Well, I don't like it as well but how to deal with this real-real-world problem otherwise? > -Sean > > Moreno Carullo wrote: > > Hi all, > > I'm going to implement a way to start a search on Lucene with a > > user-defined timeout. > > This is really useful when you have to deal with very strict SLA, and for > > what I have seen from the source code, this feature is still missing. > > > > Has anyone start doing something similar? > > Does my future patch has any chance to be included in standard lucene? > > > > Thanks > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]