[jira] [Created] (YARN-9073) GPU/FPGA whitelist configuration in container-executor.cfg won't work when yarn-site.xml's allowed devices doesn't align with it

["Zhankun Tang (JIRA)" <ji...@apache.org>]

Zhankun Tang created YARN-9073:
----------------------------------

             Summary: GPU/FPGA whitelist configuration in container-executor.cfg won't work when yarn-site.xml's allowed devices doesn't align with it
                 Key: YARN-9073
                 URL: https://issues.apache.org/jira/browse/YARN-9073
             Project: Hadoop YARN
          Issue Type: Bug
            Reporter: Zhankun Tang
            Assignee: Zhankun Tang


The current GPU/FPGA behavior may has an issue when c-g.cfg doesn't align with yarn-site.xml. Take GPU for instance:

One host has 1,2,3,4,5. And "GPU.allowed = 1,2,3" configured in c-e.cfg. But yarn-site.xml configured auto which means 1,2,3,4,5.

And one application request 4 GPU, the scheduler allocated 1,2,4,5. So --excluded-gpus is "3". And c-e will check that 3 is in allowed list(1,2,3) and then only deny 3 in cgroups.

In this case, c-e's allowed-list (1,2,3) doesn't work because the application can access 4 and 5.



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

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