[torqueusers] checkjob shows RM failure and Invalid Credential

Eva Hocks hocks at sdsc.edu
Wed Oct 2 11:16:46 MDT 2013



My first guess would be the /etc/passwd file not synced on all
your compute nodes.

-Eva


On Wed, 2 Oct 2013, Charles Johnson wrote:

> I have a user's job not able to start up though torque has place the job
> and checkjob shows the job in the running state. Checkjob shows this
> message:
>
> Message[0] cannot start job 626200 - RM failure, rc: 15021, msg:
> 'Invalid credential MSG=Users do not match'
>
> I have not seen this error message before. Any ideas on what it means?
> If the resource manager has failed on one of the nodes --there are 200
> nodes in this job-- ideas on how to track it down?
>
> What does the "'Invalid credential MSG=Users do not match" mean?
>
> Charles
>
>



More information about the torqueusers mailing list