[
https://issues.apache.org/jira/browse/HDDS-797?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16673773#comment-16673773
]
Hanisha Koneru commented on HDDS-797:
-------------------------------------
This issue happens because the DN queues multiple GetVersion tasks while
waiting to connect with SCM. When SCM eventually comes up, all the GetVersion
tasks still in the queue (which have not timed out retrying) execute
successfully and update the endpoint state of the DN state machine.
The first GetVersion task to successfully get the version from SCM updates the
state from GETVERSION to HEARTBEAT. The second GetVersion task then updates
this state from HEARTBEAT to SHUTDOWN.
Once DN goes into HEARTBEAT state, it should not execute the GetVersion tasks.
> If DN is started before SCM, it does not register
> -------------------------------------------------
>
> Key: HDDS-797
> URL: https://issues.apache.org/jira/browse/HDDS-797
> Project: Hadoop Distributed Data Store
> Issue Type: Bug
> Reporter: Mukul Kumar Singh
> Assignee: Hanisha Koneru
> Priority: Blocker
> Attachments: HDDS-797.001.patch
>
>
> If a DN is started before SCM, it does not register with the SCM. DNs keep
> trying to connect with the SCM and once SCM is up, the DN services are
> shutdown instead of registering with SCM.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
