[
https://issues.apache.org/jira/browse/HIVE-6635?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13932695#comment-13932695
]
Ashutosh Chauhan commented on HIVE-6635:
----------------------------------------
This will double # of threads for HS2 and worse these threads are not coming
from shared threadpool of HS2. How about putting heartbeat while we are waiting
for MR job to finish. There is a while loop waiting polling for MR job status.
I think we can put heartbeating there. This is in
HadoopJobExecHelper::progress()
> Heartbeats are not being sent when DbLockMgr is used and an operation holds
> locks
> ---------------------------------------------------------------------------------
>
> Key: HIVE-6635
> URL: https://issues.apache.org/jira/browse/HIVE-6635
> Project: Hive
> Issue Type: Bug
> Components: Locking
> Affects Versions: 0.13.0
> Reporter: Alan Gates
> Assignee: Alan Gates
> Fix For: 0.13.0
>
> Attachments: HIVE-6635.patch
>
>
> The new DbLockManager depends on heartbeats from the client in order to
> determine that a lock has not timed out. The client is not currently sending
> those heartbeats.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
