[slurm-users] AllocNodes on partition no longer working

[Christopher Samuel]

On 8/14/19 10:46 AM, Sajdak, Doris wrote:

> We upgraded from version 18.08.4 to 19.05.1-2 today and are suddenly 
> getting a permission denied error on partitions where we have AllocNodes 
> set.  If we remove the AllocNodes constraint, the job submits 
> successfully but then users can submit from anywhere which is not what 
> we want.  Has anyone else seen this problem?
> 
> sbatch: error: Batch job submission failed: Access/permission denied

It's working here - though we got caught out with that because the IP 
address was being resolved to the FQDN of the node and not the short 
name we had in our config file.

To see what your system is resolving the IP address to now use scontrol 
to set your debug level to "debug2" and see what it reports when the 
test fails (it would be nice if Slurm actually logged that as an error).

All the best,
Chris
-- 
   Chris Samuel  :  http://www.csamuel.org/  :  Berkeley, CA, USA