[
https://issues.apache.org/jira/browse/SOLR-686?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12621160#action_12621160
]
Yonik Seeley commented on SOLR-686:
-----------------------------------
I guess because I don't see any use-case or reason for using NoLockFactory
other than some kind of testing.
If an index is truly read-only then a real lock factory won't have any overhead.
But a real lock factory is a nice backup in the event that the configuration is
incorrect and it becomes writable. People may also mistakenly believe that
using "none" is more efficient somehow on read-only indexes.
So since I'd never recommend using "none" in any scenario, it doesn't seem like
it should even be listed as an option in the example schema. One could make
the case that "none" should be removed altogether, but I didn't go quite that
far.
> single lock factory overwrites previous
> ---------------------------------------
>
> Key: SOLR-686
> URL: https://issues.apache.org/jira/browse/SOLR-686
> Project: Solr
> Issue Type: Bug
> Components: update
> Reporter: Yonik Seeley
> Assignee: Yonik Seeley
> Fix For: 1.3
>
> Attachments: SOLR-686.patch
>
>
> On a core reload, the Directory is retrieved and a new single lock factory is
> set, effectively removing all previous locks.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
