[slurm-users] New Bright Cluster Slurm issue for AD users

[Loris Bennett]

Yugendra Guvvala <yguvvala at cambridgecomputer.com> writes:

>> On Feb 13, 2019, at 1:50 AM, Loris Bennett <loris.bennett at fu-berlin.de> wrote:
>> 
>> Yugendra Guvvala <yguvvala at cambridgecomputer.com> writes:
>> 
>>> Hi, 
>>> 
>>> We are bringing a new cluster online. We installed SLURM through Bright Cluster Manager how ever we are running into a issue here. 
>>> 
>>> We are able to run jobs as root user and users created using bright cluster (cmsh commands). How ever we use AD authentication for all our users and when we try to
>>> submit jobs to slurm using AD users we are getting following error message. 
>>> 
>>> srun: fatal: Invalid user id: 10952
>>> srun: fatal: Invalid user id: 10952
>>> srun: error: cnode001: task 0: Exited with exit code 1
>>> 
>>> Attached is the slurm.con file for reference. Please let us know if you have any insight into this. 
>>> 
>>> Thanks, 
>>> Yugi
>> 
>> Are your AD users visible in the output of the following command:
>> 
>>  sacctmgr list assoc
>> 
>> ?  If not, you need to set them up, also using 'sacctmgr'.  The cmsh
>> command for adding users maybe doing this in the background (although in
>> my experience with BCM 5, this did not happen).
>> 
> Hi Loris,
>
> Thank you, this is strange. Is there a way to integrate AD
> authentication with SLURM or Munge. Or allow all users who login to
> run jobs with out any restrictions .

I'm not sure whether this is strange - it allows you to set up a
hierarchy of groups/projects/whatever, regardless of what is in your
AD.

In any case, normally you have to set

  AccountingStorageEnforce=associations

to prevent users from starting jobs if they do not exist in the Slurm
accounting database.  As you don't seem to have done that, I can't see
what the problem is.  Does

  getent passwd 10952

return the user information properly?

Cheers,

Loris

-- 
Dr. Loris Bennett (Mr.)
ZEDAT, Freie Universität Berlin         Email loris.bennett at fu-berlin.de