hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Lars Hofhansl (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-9869) Optimize HConnectionManager#getCachedLocation
Date Mon, 18 Nov 2013 02:11:20 GMT

    [ https://issues.apache.org/jira/browse/HBASE-9869?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13825038#comment-13825038
] 

Lars Hofhansl commented on HBASE-9869:
--------------------------------------

I have suggested changing this in the past, but the consensus used to be that we do not want
to remove the soft value reference map (which would allow the GC to clean out entries from
the map when memory is tight).

I think soft references for caching are an anti-pattern (we're short of memory so we clear
the cache in order to do more work if we need the cached data again)... But I wonder whether
something has changed in the past year that makes that OK now.

Anyway, +1 from my side.

> Optimize HConnectionManager#getCachedLocation
> ---------------------------------------------
>
>                 Key: HBASE-9869
>                 URL: https://issues.apache.org/jira/browse/HBASE-9869
>             Project: HBase
>          Issue Type: Bug
>          Components: Client
>    Affects Versions: 0.98.0, 0.96.0
>            Reporter: Nicolas Liochon
>            Assignee: Nicolas Liochon
>             Fix For: 0.98.0, 0.96.1
>
>         Attachments: 6869.v4.patch, 9869.v1.patch, 9869.v1.patch, 9869.v2.patch
>
>
> It javadoc says: "TODO: This method during writing consumes 15% of CPU doing lookup".
This is still true, says Yourkit. With 0.96, we also spend more time in these methods. We
retry more, and the AsyncProcess calls it in parallel.
> I don't have the patch for this yet, but I will spend some time on it.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

Mime
View raw message