hive-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alan Gates (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HIVE-12529) HiveTxnManager.acquireLocks() should not block forever
Date Wed, 02 Dec 2015 18:23:10 GMT

    [ https://issues.apache.org/jira/browse/HIVE-12529?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15036308#comment-15036308
] 

Alan Gates commented on HIVE-12529:
-----------------------------------

I don't understand what you're trying to do here.  It looks like you want to unlock the waiting
locks without rolling back the transaction.  Is that true?  If so, why?  Allowing this violates
a lot of assumptions in the code and I'm afraid we'll cause a lot of bugs.  Why not roll back
the transaction at that point?


> HiveTxnManager.acquireLocks() should not block forever
> ------------------------------------------------------
>
>                 Key: HIVE-12529
>                 URL: https://issues.apache.org/jira/browse/HIVE-12529
>             Project: Hive
>          Issue Type: Bug
>          Components: Transactions
>    Affects Versions: 1.0.0
>            Reporter: Eugene Koifman
>            Assignee: Eugene Koifman
>         Attachments: HIVE-12529.2.patch, HIVE-12529.4.patch, HIVE-12529.patch
>
>
> Currently, in DbTxnManager this method will block until all competing locks have gone
away.
> This is not appropriate for all clients.  There should be a way to specify a max-wait-time.
> It will throw an exception on timeout (given how current method signature is written).



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message