[ 
https://issues.apache.org/jira/browse/HIVE-19340?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16456868#comment-16456868
 ] 

Sergey Shelukhin edited comment on HIVE-19340 at 4/27/18 6:25 PM:
------------------------------------------------------------------

Never aborting a transaction seems dangerous. 
If it does get stuck forever, what is the user supposed to do with it? 
And if the user aborts it manually, it's the same as timeout, just aggravating 
to the user. You still have to handle when user aborts it.

If it's ok to get stuck for a long time but not forever, why not just increase 
the heartbeat timeout for it?


was (Author: sershe):
Never aborting a transaction seems dangerous. 
If it does get stuck forever, what is the user supposed to do with it? 
And if the user aborts it manually, it's the same as timeout, just aggravating 
to the user. You still have to handle when user aborts it.

If it's ok to get stuck for a long time, why not just increase the heartbeat 
timeout for it?

> Disable timeout of transactions opened by replication task at target cluster
> ----------------------------------------------------------------------------
>
>                 Key: HIVE-19340
>                 URL: https://issues.apache.org/jira/browse/HIVE-19340
>             Project: Hive
>          Issue Type: Sub-task
>          Components: repl, Transactions
>    Affects Versions: 3.0.0
>            Reporter: mahesh kumar behera
>            Assignee: mahesh kumar behera
>            Priority: Major
>              Labels: ACID, DR, pull-request-available, replication
>             Fix For: 3.0.0
>
>         Attachments: HIVE-19340.01.patch
>
>
> The transactions opened by applying EVENT_OPEN_TXN should never be aborted 
> automatically due to time-out. Aborting of transaction started by replication 
> task may leads to inconsistent state at target which needs additional 
> overhead to clean-up. So, it is proposed to mark the transactions opened by 
> replication task as special ones and shouldn't be aborted if heart beat is 
> lost. This helps to ensure all ABORT and COMMIT events will always find the 
> corresponding txn at target to operate.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to