[
https://issues.apache.org/jira/browse/SOLR-571?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12859050#action_12859050
]
Tomás Fernández Löbbe commented on SOLR-571:
--------------------------------------------
I don't understand why it could take longer to commit than when using an
absolute value instead of a percentage.
The number of items sent from the old cache to the new cache will depend on the
size of the old cache and on the configured percentage. If someone uses a big
percentage (like 75% or 100%) on a big cache, it might take long, but it will
be pretty much the same of using a very big number on autowarmCount attribute
on the same cache.
Am I missing something? I can't see a problem here.
> LRUCache autowarmCount should support percentages
> -------------------------------------------------
>
> Key: SOLR-571
> URL: https://issues.apache.org/jira/browse/SOLR-571
> Project: Solr
> Issue Type: Improvement
> Reporter: Hoss Man
>
> a recent thread reminded me of an old idea that never got implemented...
> the autowarmCount for LRUCaches should support "percentages" which get
> evaluated relative the current size of the cache when warming happens.
> in this way, a Solr instance might be configured with autowarmCount="50%" to
> autowarm the top half of the queries ... a master machine might be configured
> this way to give some autowarming if it is inadvertently being queried, but
> things won't be maintained in the caches in perpetuity ... each subsequent
> cache instance would have fewer and fewer of the "old" queries warmed
> automatically.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
