[
https://issues.apache.org/jira/browse/SOLR-667?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12643520#action_12643520
]
Todd Feak commented on SOLR-667:
--------------------------------
Huge thanks on this one. This was one of the bottlenecks I've seen previously.
For apples to apples load tests, this more then doubled my overall throughput.
I do notice a sort of "pulsing" in the responses. It appears that everything
flies along, but on occasion everything piles up for a second, then starts
going again. This leads to a few response times that are over 1 second, but the
average is way down closer to 20ms. Is this the cleanup involved when hitting a
high-water mark?
Overall, it's a huge improvement.
> Alternate LRUCache implementation
> ---------------------------------
>
> Key: SOLR-667
> URL: https://issues.apache.org/jira/browse/SOLR-667
> Project: Solr
> Issue Type: New Feature
> Components: search
> Affects Versions: 1.3
> Reporter: Noble Paul
> Assignee: Shalin Shekhar Mangar
> Fix For: 1.4
>
> Attachments: ConcurrentLRUCache.java, ConcurrentLRUCache.java,
> ConcurrentLRUCache.java, 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.
