hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Nanda kumar (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDDS-428) OzoneManager lock optimization
Date Wed, 12 Sep 2018 08:11:00 GMT

    [ https://issues.apache.org/jira/browse/HDDS-428?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16611735#comment-16611735

Nanda kumar commented on HDDS-428:

Thanks [~anu] for the review. Addressed the comments in patch v01.
{quote}Should we also hold volumeLock?
While deleting a bucket, we never touch {{volume table}}, I don't see why we should acquire
a volume lock as it doesn't change or do anything to the volume table.
{quote}Looks like we are holding volume lock? do we need this ?
Unless there is a rogue client which calls {{addBlock}} from multiple threads at the same
time, we don't need this lock. In such cases, the client will be in much more trouble. I don't
think we should handle those cases.
Removed the lock from {{KeyManagerImpl#allocateBlock}}

> OzoneManager lock optimization
> ------------------------------
>                 Key: HDDS-428
>                 URL: https://issues.apache.org/jira/browse/HDDS-428
>             Project: Hadoop Distributed Data Store
>          Issue Type: Improvement
>          Components: OM
>            Reporter: Nanda kumar
>            Assignee: Nanda kumar
>            Priority: Major
>             Fix For: 0.2.1
>         Attachments: HDDS-428.000.patch, HDDS-428.001.patch
> Currently, {{OzoneManager}} uses a single lock for everything which impacts the performance.
We can introduce a separate lock for each resource like User/Volume/Bucket which will give
us a performance boost.

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org

View raw message