[
https://issues.apache.org/jira/browse/YARN-6002?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15748843#comment-15748843
]
Raul Gutierrez Segales commented on YARN-6002:
----------------------------------------------
The provided implementation is actually different than what is in trunk:
https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/webapp/NMWebServices.java#L291
We might as well backport that.
> Range support when serving a container log
> ------------------------------------------
>
> Key: YARN-6002
> URL: https://issues.apache.org/jira/browse/YARN-6002
> Project: Hadoop YARN
> Issue Type: Bug
> Components: nodemanager
> Affects Versions: 2.7.3
> Reporter: Raul Gutierrez Segales
> Priority: Minor
> Fix For: 2.7.4
>
> Attachments: YARN-6002-branch-2.7.001.patch
>
>
> Currently, when we access a container's logfile (via
> /ws/v1/node/containerlogs/) we can only get the full content, not a specified
> range.
> Support for the Range: header would improve this, special when dealing with
> big files that need to be tailed or streamed, i.e.:
> {code}
> curl -H 'Range: bytes=10240' \
> http://node:8042/ws/v1/node/containerlogs/container_XXX/stdout
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
