incubator-blur-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Patrick Hunt (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (BLUR-5) Write through caching for the BlockCache
Date Wed, 24 Oct 2012 21:20:13 GMT

    [ https://issues.apache.org/jira/browse/BLUR-5?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13483602#comment-13483602
] 

Patrick Hunt commented on BLUR-5:
---------------------------------

What other changes should I make before committing this? I'm thinking the following two:

1) make the write and read caching behavior configurable - ie individually on or off for an
indexdir
2) Use the context as suggested by Doug to control default read/write cache behavior "cache
i/o that occurs in a READ or FLUSH context, but only cache MERGE-related i/o when the estimatedMergeBytes
is substantially smaller than the cache size". I'd also make these configurable so that it
would  be easy to experiment.

Other than that the full unit test suite is now passing for me so I should be pretty close
to being able to commit this. What do you think?
                
> Write through caching for the BlockCache
> ----------------------------------------
>
>                 Key: BLUR-5
>                 URL: https://issues.apache.org/jira/browse/BLUR-5
>             Project: Apache Blur
>          Issue Type: Improvement
>            Reporter: Aaron McCurry
>            Assignee: Patrick Hunt
>         Attachments: BLUR-5.patch, BLUR-5.patch
>
>
> This will allow for better NRT update performance because the writer will not have to
read the NRT segments from HDFS.

--
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

Mime
View raw message