hi,
we observed a strange behavior of pam_slurm_adopt regarding the
involved cgroups:
when we start a shell as a new Slurm job using "srun", the process has
freezer, cpuset and memory cgroups setup as e.g.
"/slurm/uid_5001/job_410318/step_0". that's good!
however, another shell started by
hi,
thank you patrick, thank you kilian for identifying a systemd issue here!
for a quick test, we disabled and masked systemd-logind. the "memory"
cgroup now works as expected. great!
we're now watching out for any unwanted side effects.
cheers,
christian
On 08/22/2018 07:02 PM, Patrick
hi chris,
we're using the same distro.
regards,
christian
On 08/23/2018 01:02 PM, Chris Samuel wrote:
I don't see that on our CentOS 7.5 system, which distro are you using?
<>
