[ 
https://issues.apache.org/jira/browse/HDDS-238?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16536417#comment-16536417
 ] 

Ajay Kumar commented on HDDS-238:
---------------------------------

[~msingh] thanks for working on this. Patch LGTM. Few suggestions:
 * ContainerStateMap
 ** L88: typo for space. May be we can rephrase that line to something like.
{code}This map will be queried for the list of open containers on a particular 
pipeline and issue a close onĀ correspondingĀ containers in case of following 
events:
1. Dead datanode.
2. Datanode out of space.
3. Volume loss or volume out of space.
{code}
 ** addContainer L144: Since open container is also called from 
ContainerStateManager#loadExistingContainers there might be container which are 
not open and might not have pipeline id as its optional field in protobuf. We 
should check for Container state and pipeline id.
* Node2PipelineMap
** L57: typo (missing "contains")
** updateDatanodeMap L96: BiFunction argument doesn't seems correct. I think 
this will remove any existing values for given dnId.
** addPipeline/updateDatanodeMap: IMO we have to synchronize hashset before we 
modify its state.
** getPipelines: Wouldn't returning unmodifiable set from get operation will be 
bit less expensive?
* TestNode2PipelineMap: Shall we update testcase to test update op of  
Node2PipelineMap and openPipelineMap?

> Add Node2Pipeline Map in SCM to track ratis/standalone pipelines.
> -----------------------------------------------------------------
>
>                 Key: HDDS-238
>                 URL: https://issues.apache.org/jira/browse/HDDS-238
>             Project: Hadoop Distributed Data Store
>          Issue Type: Bug
>          Components: SCM
>    Affects Versions: 0.2.1
>            Reporter: Mukul Kumar Singh
>            Assignee: Mukul Kumar Singh
>            Priority: Major
>             Fix For: 0.2.1
>
>         Attachments: HDDS-238.001.patch
>
>
> This jira proposes to add a Node2Pipeline map which can be used to during 
> pipeline failure to identify a pipeline for a corresponding failed datanode.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to