lucene-solr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Todd Feak (JIRA)" <>
Subject [jira] Commented: (SOLR-667) Alternate LRUCache implementation
Date Thu, 30 Oct 2008 20:38:46 GMT


Todd Feak commented on SOLR-667:

I ran with a profiler and I'm not seeing any bursts of garbage collection. It's at a steady
~2%, with no major collections occurring (which is great!). However, the use of the profiler
also slows things down about 10-20 % which seems to be enough that the pulsing goes away.
I believe the pulsing may be some sort of limit of the testing I'm doing on a limited local
environment. I'll include the patch in my current Solr app and do a more accurate comparison
with our production level load tests to see if it still exists. Though it may be a while before
I can provide feedback from that one, as getting machines allocated for the heavy load testing
can take a bit.

As I said before, this is a huge improvement. Thanks for all the work on this one.

> Alternate LRUCache implementation
> ---------------------------------
>                 Key: SOLR-667
>                 URL:
>             Project: Solr
>          Issue Type: New Feature
>          Components: search
>    Affects Versions: 1.3
>            Reporter: Noble Paul
>            Assignee: Shalin Shekhar Mangar
>             Fix For: 1.4
>         Attachments:,,,
SOLR-667-updates.patch, SOLR-667.patch, SOLR-667.patch, SOLR-667.patch, SOLR-667.patch, SOLR-667.patch,
SOLR-667.patch, SOLR-667.patch
> The only available SolrCache i.e LRUCache is based on _LinkedHashMap_ which has _get()_
also synchronized. This can cause severe bottlenecks for faceted search. Any alternate implementation
which can be faster/better must be considered. 

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message