hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ajay Kumar (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDDS-297) Add pipeline actions in Ozone
Date Tue, 21 Aug 2018 19:34:00 GMT

    [ https://issues.apache.org/jira/browse/HDDS-297?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16587908#comment-16587908
] 

Ajay Kumar commented on HDDS-297:
---------------------------------

[~msingh] thanks for updating the patch. LGTM. Few minor comments and questions:
* ContainerMapping#handlePipelineClose log pipeline id if its not found?
* ContainerStateMachine L119 rename server to ratisServer
* HDDSConfigKeys Just curious about 20 limit. For a medium to big cluster this may be too
small?
* RatisHelper L58. Is this "_" in peerId generated by Ratis API? If yes, than may be we should
replace it with an internal config constant. (To avoid any potential breakage by ratis in
future)
* StateContext L275 Shall we add unordered() before limit? This is what javadoc says about
limit.
{code}Using an unordered
     * stream source (such as {@link #generate(Supplier)}) or removing the
     * ordering constraint with {@link #unordered()} may result in significant
     * speedups of {@code limit()} in parallel pipelines, if the semantics of
     * your situation permit.  If consistency with encounter order is required,
     * and you are experiencing poor performance or memory utilization with
     * {@code limit()} in parallel pipelines, switching to sequential execution
     * with {@link #sequential()} may improve performance.{code}
     
* StorageContainerDatanodeProtocol.proto Could you please share why ClosePipelineInfo in PipelineAction
is optional? In PipelineEventHandler that seems to be required field.
* PipelineEventHandler: Rename class to PipelineActionEventHandler?
* TestNodeFailure#testPipelineFail Shall we assert that ratisContainer1 pipeline is Open before
we shutdown the datanode in that pipeline? Also shall we test failure for leader and follower
separately?

> Add pipeline actions in Ozone
> -----------------------------
>
>                 Key: HDDS-297
>                 URL: https://issues.apache.org/jira/browse/HDDS-297
>             Project: Hadoop Distributed Data Store
>          Issue Type: Bug
>          Components: SCM
>            Reporter: Mukul Kumar Singh
>            Assignee: Mukul Kumar Singh
>            Priority: Major
>             Fix For: 0.2.1
>
>         Attachments: HDDS-297.001.patch, HDDS-297.002.patch, HDDS-297.003.patch
>
>
> Pipeline in Ozone are created out of a group of nodes depending upon the replication
factor and type. These pipeline provide a transport protocol for data transfer.
> Inorder to detect any failure of pipeline, SCM should receive pipeline reports from Datanodes
and process it to identify various raft rings.



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

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


Mime
View raw message