[jira] [Commented] (YARN-9561) Add C changes for the new RuncContainerRuntime

["Eric Yang (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/YARN-9561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16921670#comment-16921670 ] 

Eric Yang commented on YARN-9561:
---------------------------------

[~ebadger] {quote}
I don't think this should be added in the code that was created by this patch. The functions that are called from within setup_container_paths() are the ones that should give more fine-grained error logging. Because at the setup_container_paths() level I have no idea which directory failed to be created. I would only be able to give a large list. So I think this is an improvement for the underlying functions, but isn't relevant to this specific patch.
{quote}

You are right that create_script_paths already provide verbose message of what failed.  I was thinking that we could print the returned parameters, but that would be redundant to create_script_paths function.

{quote}The 16k size is just the initial size to create upperdir=%s,workdir=%s,lowerdir=. Upperdir and workdir can only be single directories, so the only way the 16k limit would be exceeded is if these directory paths were absolutely gigantic. Inside of strbuf_append_fmt(){quote}

Strbuf_append_fmt is also used for recording layer information into sb variable.  Layers information may exceed 16k boundary.  I am not as confident with the buffer resize in current implementation of strbuf_append_fmt and pointer tracking is bug free.  I think there is a bug in reprint the remaining buffer.  Please provide test case for this function.  Thanks

> Add C changes for the new RuncContainerRuntime
> ----------------------------------------------
>
>                 Key: YARN-9561
>                 URL: https://issues.apache.org/jira/browse/YARN-9561
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>            Reporter: Eric Badger
>            Assignee: Eric Badger
>            Priority: Major
>         Attachments: YARN-9561.001.patch, YARN-9561.002.patch, YARN-9561.003.patch, YARN-9561.004.patch
>
>
> This JIRA will be used to add the C changes to the container-executor native binary that are necessary for the new RuncContainerRuntime. There should be no changes to existing code paths. 



--
This message was sent by Atlassian Jira
(v8.3.2#803003)

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