[
https://issues.apache.org/jira/browse/HDDS-17446?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ASF GitHub Bot updated HDDS-17446:
----------------------------------
Labels: pull-request-available (was: )
> Avoid copying the datanode container set twice in NodeStateMap.updateNode
> -------------------------------------------------------------------------
>
> Key: HDDS-17446
> URL: https://issues.apache.org/jira/browse/HDDS-17446
> Project: Apache Ozone
> Issue Type: Improvement
> Reporter: Andy Huang
> Assignee: Andy Huang
> Priority: Minor
> Labels: pull-request-available
>
> NodeStateMap.updateNode preserves a datanode's containers by passing
> oldEntry.copyContainers() to the DatanodeEntry constructor. copyContainers()
> creates a new HashSet, and the constructor then copies that set again with
> addAll. This allocates an unnecessary intermediate set and traverses the
> containers twice while holding the NodeStateMap write lock.
> Pass the old DatanodeEntry to the constructor and copy old.containers
> directly into the new entry's container set. This preserves an independent
> set with a single copy.
> Follow-up to HDDS-16632, identified in review:
> https://github.com/apache/ozone/pull/11687#discussion_r4357379505
> Validation: TestSCMNodeManager#testRegisterKeepsContainersOnNodeUpdate passes
> with this change (1 test, no failures or errors).
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
