Cyclic distribution problem

[Tsai Li Ming]

Hi,

I’m running 2 slurmds on a single host (built with --enable-multiple-slurmd). The total cpus are divided equally among the 2 nodes.

I’m trying to test the distribution modes=block/cyclic but the tasks are always allocated on the first node unless I use "--ntasks-per-node=1"

$ srun -n2 --distribution=block/cyclic sleep 100

I’m using:
SelectType=select/cons_res
SelectTypeParameters=CR_CPU_MEMORY
NodeName=ltsai-dev-rhel7-1 NodeHostname=ltsai-dev-rhel7 Port=17001 Sockets=1 CoresPerSocket=2 ThreadsPerCore=1 RealMemory=1841 State=UNKNOWN
NodeName=ltsai-dev-rhel7-2 NodeHostname=ltsai-dev-rhel7 Port=17002 Sockets=1 CoresPerSocket=2 ThreadsPerCore=1 RealMemory=1841 State=UNKNOWN
PartitionName=compute Nodes=ltsai-dev-rhel7-[1-2] Default=YES MaxTime=INFINITE State=UP

Did I misconfigure something?

Thanks!=