You are viewing a plain text version of this content. The canonical link for it is here.

Posted to mapreduce-issues@hadoop.apache.org by "Ted Yu (JIRA)" <ji...@apache.org> on 2014/04/03 00:43:18 UTC

[jira] [Commented] (MAPREDUCE-5819) Binary token merge should be done once in TokenCache#obtainTokensForNamenodesInternal()

    [ https://issues.apache.org/jira/browse/MAPREDUCE-5819?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13958275#comment-13958275 ] 

Ted Yu commented on MAPREDUCE-5819:
-----------------------------------

TestTokenCache and TestBinaryTokenFile pass based on patch.

> Binary token merge should be done once in TokenCache#obtainTokensForNamenodesInternal()
> ---------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-5819
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5819
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>            Reporter: Ted Yu
>            Assignee: Ted Yu
>            Priority: Minor
>         Attachments: mapreduce-5819-v1.txt
>
>
> Currently mergeBinaryTokens() is called by every invocation of obtainTokensForNamenodesInternal(FileSystem, Credentials, Configuration) in the loop of obtainTokensForNamenodesInternal(Credentials, Path[], Configuration).
> This can be simplified so that mergeBinaryTokens() is called only once in obtainTokensForNamenodesInternal(Credentials, Path[], Configuration).



--
This message was sent by Atlassian JIRA
(v6.2#6252)