[slurm-users] CPU allocation within a node is not cyclic

[崔灏 (CUI Hao)]

Hi,

According to https://slurm.schedmd.com/cpu_management.html,
> The default allocation method within a node is cyclic allocation (allocate available CPUs in a round-robin fashion across the sockets within a node).

Not a native English speaker. I think the sentense means that: if a
job allocates 2*n cores (with -c 2*n) on a node with two physical CPUs
(sockets), n cores will be allocated on each sockets by default.

However, on our cluster core allocation is not evenly across different
sockets. I cannot find the problem in our configs. Explicit setting
with srun --cpu-bind or --distribution cyclic doesn't work, either.

For example: `srun -p cpu1 -c 8` results continuous allocation on a
single socket. `srun -p cpu1 -c 14` results allocation of 2 cores + 12
cores.

Here is our config and Slurm version is 18.08.0:
https://gist.github.com/cuihaoleo/6086cb64f078d86af844b560c5d8228a

I really hope to get the default cylic allocation behaviour, so that
CPU jobs won't take up cores assigned to GPU devices. Any idea what's
wrong with our configs?

--
崔灏 / CUI Hao
Homepage: i-yu.me
Twitter: @cuihaoleo