You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@sling.apache.org by "Chetan Mehrotra (JIRA)" <ji...@apache.org> on 2015/07/20 12:31:04 UTC
[jira] [Commented] (SLING-4891) Improve MapEntries to cache
searched vanity paths
[ https://issues.apache.org/jira/browse/SLING-4891?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14633394#comment-14633394 ]
Chetan Mehrotra commented on SLING-4891:
----------------------------------------
[~asanso] Another aspect which we should look for is support for negative cache. This would be useful for case where a given url is considered by bloom filter as a vanity path (false positive) but is actually not a vanity url. For such a case current logic would end up firing a new query always. For such cases have a negative cache.
Now to determine if such cases are common we would need some JMX stats around how effective is vanity path logic in presence of bloom filter
> Improve MapEntries to cache searched vanity paths
> --------------------------------------------------
>
> Key: SLING-4891
> URL: https://issues.apache.org/jira/browse/SLING-4891
> Project: Sling
> Issue Type: Improvement
> Components: ResourceResolver
> Reporter: Antonio Sanso
> Assignee: Antonio Sanso
>
> At the moment {{PROP_MAX_CACHED_VANITY_PATHS}} limits the number of in memory cached vanity paths in order to have a faster startup.
> I'd propose to slightly change the semantic of the limit of the cache .
> We can indeed limit the cache only for startup and when a vanity path query is executed and results are returned those can be added to the cache (and not thrown away as is now).
> If this same entry will be again requested it will be delivered from the cache, so fast.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)