lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ian Lea" <ian....@gmail.com>
Subject Re: Cache Used by IndexReader/IndexSearcher
Date Thu, 18 Dec 2008 09:56:36 GMT
Hi


Are all the queries broadly similar or are the later ones more
complex?  What happens if you switch the order and run the later
queries first?  Any complications like sorting?  Has your jvm got
enough memory?

There is no IndexSearcher cache that you can increase.


--
Ian.

On Wed, Dec 17, 2008 at 12:19 AM, Sangrish <sidangrish@gmail.com> wrote:
>
> Hi All,
>
>          I have a 50 GB index of about 40 million documents. I need to
> query it
> around 40,000 times(different queries) one by one. I saw that the query
> times are negligible for the
> first, say 25,000 queries, but it degrades later on. For example, the time
> for 200 sequential queries changes from 0.009 secs to 0.06 secs. Any
> possible reasons?
>
>   I read somewhere that IndexSearcher reads in a cache when its
> initialized. I could not find more on this.
> Can I increase this cache size to get an elongated performance? If yes, how
> do I play with it?
>
> Thanks
> Siddharth
>
>
> --
> View this message in context: http://www.nabble.com/Cache-Used-by-IndexReader-IndexSearcher-tp21044811p21044811.html
> Sent from the Lucene - Java Users mailing list archive at Nabble.com.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org


Mime
View raw message