[
https://issues.apache.org/jira/browse/HDDS-308?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16571988#comment-16571988
]
Lokesh Jain commented on HDDS-308:
----------------------------------
[~msingh] Thanks for reviewing the patch! v5 patch addresses your comments. I
have also made a change in DeletedBlockLogImpl:258 where I check for
replication factor. This change is done so that if a close container is present
in two datanodes when the replication factor is three we do not remove the
deletedBlock entries from the log. Maybe I should keep a check on number of dns
received in
{code:java}
containerManager.getContainerWithPipeline(containerId)
{code}
call. If the number of dns received is less than the replication factor we
should not proceed with committing a txn?
> SCM should identify a container with pending deletes using container reports
> ----------------------------------------------------------------------------
>
> Key: HDDS-308
> URL: https://issues.apache.org/jira/browse/HDDS-308
> Project: Hadoop Distributed Data Store
> Issue Type: Bug
> Components: SCM
> Reporter: Lokesh Jain
> Assignee: Lokesh Jain
> Priority: Major
> Fix For: 0.2.1
>
> Attachments: HDDS-308.001.patch, HDDS-308.002.patch,
> HDDS-308.003.patch, HDDS-308.004.patch, HDDS-308.005.patch
>
>
> SCM should fire an event when it finds using container report that a
> container's deleteTransactionID does not match SCM's deleteTransactionId.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
