[
https://issues.apache.org/jira/browse/LUCENE-5038?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13679830#comment-13679830
]
Simon Willnauer commented on LUCENE-5038:
-----------------------------------------
hey maybe we should not allow to set it at all and just write CFS by default? ;)
I really think this is not that much of a deal here, I kind of like the flush
policy idea though.
bq. So now we have 3 settings: mp.setUseCFS(), mp.setNoCFSRatio() and
iwc.setUseCFSOnWrite. That's too much, no?
Thinking out loud... I couldn't agree more, I think mp.setUseCFS() should go
away entirely and we should only use setNonCFSRatio() to control for merges.
0.0 would mean no CFS in that case. on IWC we just use the setting from the MP
as the default ie. if you disable it we also disable it in the IWC unless it's
set explicitly via a setter. That means we are back to 2 settings, MP implies
the setting on IWC...
Yet, the MP settings are impl details and not part of the interface at all
(which is ok I think?) would we awesome if we could clean this up even further
in the future :)
> Don't call MergePolicy / IndexWriter during DWPT Flush
> ------------------------------------------------------
>
> Key: LUCENE-5038
> URL: https://issues.apache.org/jira/browse/LUCENE-5038
> Project: Lucene - Core
> Issue Type: Improvement
> Components: core/index
> Affects Versions: 4.3
> Reporter: Simon Willnauer
> Assignee: Simon Willnauer
> Fix For: 4.4
>
> Attachments: LUCENE-5038.patch, LUCENE-5038.patch, LUCENE-5038.patch
>
>
> We currently consult the indexwriter -> merge policy to decide if we need to
> write CFS or not which is bad in many ways.
> - we should call mergepolicy only during merges
> - we should never sync on IW during DWPT flush
> - we should be able to make the decision if we need to write CFS or not
> before flush, ie. we could write parts of the flush directly to CFS or even
> start writing stored fields directly.
> - in the NRT case it might make sense to write all flushes to CFS to minimize
> filedescriptors independent of the index size.
> I wonder if we can use a simple boolean for this in the IWC and get away with
> not consulting merge policy. This would simplify concurrency a lot here
> already.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
