[
https://issues.apache.org/jira/browse/HDDS-17?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16471100#comment-16471100
]
Xiaoyu Yao commented on HDDS-17:
--------------------------------
[~anu], thanks for working on it. The patch looks good to me overall. Here are
few minor comments:
ContainerID.java
Line 41: "positive int' should be "positive long"
Node2ContainerMap.java
Line 70: Should we use atomic APIs offered by ConcurrentHashMap like
putIfAbsent, etc. without the synchronization. This way we can take the full
advantage of it for better throughput compared with a synchronized map.
Line 99: How do we plan to share the cycle for further report (like size/stats
update) processing without looping over the containers?
Line 155: same as Line 70.
Line 159: Should we return a immutable collection here?
> Add node to container map class to simplify state in SCM
> --------------------------------------------------------
>
> Key: HDDS-17
> URL: https://issues.apache.org/jira/browse/HDDS-17
> Project: Hadoop Distributed Data Store
> Issue Type: Improvement
> Components: SCM
> Reporter: Anu Engineer
> Assignee: Anu Engineer
> Priority: Major
> Fix For: 0.2.1
>
> Attachments: HDDS-17.001.patch, HDDS-17.002.patch
>
>
> Current SCM state map is maintained in nodeStateManager. This first of
> several refactoring to make it independent and small classes.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
