hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Weiwei Yang (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HDFS-11725) Ozone: Revise create container CLI specification and implementation
Date Tue, 02 May 2017 07:12:05 GMT

     [ https://issues.apache.org/jira/browse/HDFS-11725?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

Weiwei Yang updated HDFS-11725:
    Attachment:     (was: HDFS-11725-HDFS-7240.001.patch)

> Ozone: Revise create container CLI specification and implementation
> -------------------------------------------------------------------
>                 Key: HDFS-11725
>                 URL: https://issues.apache.org/jira/browse/HDFS-11725
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: ozone
>            Reporter: Weiwei Yang
>            Assignee: Weiwei Yang
>         Attachments: HDFS-11725-HDFS-7240.001.patch
> Per [design doc|https://issues.apache.org/jira/secure/attachment/12861478/storage-container-manager-cli-v002.pdf]
in HDFS-11470
> {noformat}
> hdfs scm -container create -p <pipelineID>
> Notes : This command connects to SCM and creates a container. Once the container is created
in the SCM, the corresponding container is created at the appropriate datanode. Optional -p
allows the user to control which pipeline to use while creating this container, this is strictly
for debugging and testing.
> {noformat}
> it has 2 problems with this design, 1st it does support a container name but it is quite
useful for testing; 2nd it supports an optional option for pipeline, that is not quite necessary
right now given SCM handles the creation of the pipelines, we might want to support this later.
So proposed to revise the CLI to
> {code}
> hdfs scm -container create -c <containerName>
> {code}
> the {{-c}} option is *required*. Backend it does following steps
> # Given the container name, ask SCM where the container should be replicated to. This
returns a pipeline.
> # Communicate with each datanode in the pipeline to create the container.
> this jira is to track the work to update both the design doc as well as the implementation.

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