The problem was conflicting logic in the select/cons_res plugin. Some of the code was trying to get the job the maximum node count in the range while other logic was trying to minimize spreading out of the job across multiple switches. As you note, this problem only happens when a range of node counts is specified and the select/cons_res plugin and the topology/tree plugin and even then it is not easy to reproduce (you included all of the details below).

Quoting Martin.Perry@Bull.com:

> Certain combinations of topology configuration and srun -N option produce
> spurious job rejection with "Requested node configuration is not
> available" with select/cons_res. The following example illustrates the
> problem.
>
> [sulu] (slurm) etc> cat slurm.conf
> ...
> TopologyPlugin=topology/tree
> SelectType=select/cons_res
> SelectTypeParameters=CR_Core
> ...
>
> [sulu] (slurm) etc> cat topology.conf
> SwitchName=s1 Nodes=xna[13-26]
> SwitchName=s2 Nodes=xna[41-45]
> SwitchName=s3 Switches=s[1-2]
>
> [sulu] (slurm) etc> sinfo
> PARTITION AVAIL  TIMELIMIT  NODES  STATE NODELIST
> ...
> jkob         up   infinite      4   idle xna[14,19-20,41]
> ...
>
> [sulu] (slurm) etc> srun -N 2-4 -n 4 -p jkob hostname
> srun: Force Terminated job 79
> srun: error: Unable to allocate resources: Requested node configuration is
> not available
>
> The problem does not occur with select/linear, or topology/none, or if -N
> is omitted, or for certain other values for -N (for example, -N 4-4 and -N
> 2-3 work ok). The problem seems to be in function _eval_nodes_topo in
> src/plugins/select/cons_res/job_test.c. The srun man page states that when
> -N is used, "the job will be allocated as many nodes as possible within
> the range specified and without delaying the initiation of the job."
> Consistent with this description, the requested number of nodes in the
> above example is 4 (req_nodes=4).  However, the code that selects the
> best-fit topology switches appears to make the selection based on the
> minimum required number of nodes (min_nodes=2). It therefore selects
> switch s1.  s1 has only 3 nodes from partition jkob. Since this is fewer
> than req_nodes the job is rejected with the "node configuration" error.
>
> I'm not sure where the code is going wrong.  It could be in the
> calculation of the number of needed nodes in function _enough_nodes.  Or
> it could be in the code that initializes/updates req_nodes or rem_nodes. I
> don't feel confident that I understand the logic well enough to propose a
> fix without introducing a regression.
>
> Regards,
> Martin

When the optional max_time is not specified for --switches=count, the site
max (SchedulerParameters=max_switch_wait=seconds) is used for the job.
Based on patch from Rod Schultz.

accounting.  Work contributed by Mark Nelson.

any underlying infrastructure.

to avoid deadlock.

ionode_str

hardware in error and job running on blocks as well this fix will make it
so new blocks are formed around the bad hardware and free the old ones.

Patch from Martin Perry.

SelectType=select/cons_res
SelectTypeParameters=CR_Socket

Slurm built with ALLOCATE_FULL_SOCKET = 1

Node n8 has the following layout:
Socket 0: CPUs 0-3
Socket 1: CPUs 4-7

Without fix to _allocate_sockets (incorrect allocation for -c values of 3, 5, 6, and 7):

[sulu] (slurm) etc> srun -c1 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=4-7 Mem=0
[sulu] (slurm) etc> srun -c2 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=4-7 Mem=0
[sulu] (slurm) etc> srun -c3 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-3 Mem=0
[sulu] (slurm) etc> srun -c4 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=4-7 Mem=0
[sulu] (slurm) etc> srun -c5 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-4 Mem=0
[sulu] (slurm) etc> srun -c6 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-5 Mem=0
[sulu] (slurm) etc> srun -c7 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-6 Mem=0
[sulu] (slurm) etc> srun -c8 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-7 Mem=0

With fix to _allocate_sockets (allocation appears correct for all values of -c):

[sulu] (slurm) etc> srun -c1 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=4-7 Mem=0
[sulu] (slurm) etc> srun -c2 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=4-7 Mem=0
[sulu] (slurm) etc> srun -c3 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=4-7 Mem=0
[sulu] (slurm) etc> srun -c4 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=4-7 Mem=0
[sulu] (slurm) etc> srun -c5 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-7 Mem=0
[sulu] (slurm) etc> srun -c6 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-7 Mem=0
[sulu] (slurm) etc> srun -c7 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-7 Mem=0
[sulu] (slurm) etc> srun -c8 -m block:block --jobid 1 scontrol --details show job 1 | grep CPU_ID
     Nodes=n8 CPU_IDs=0-7 Mem=0
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A

Without this change, an assert can occur when operating bitmaps of differrent sizes

Patch by Bill Brophy, Bull.

You'll find this as commit 2079c03f58 to the chaos/slurm 2.3.3.chaos branch on March 7.  This is against SchedMD's master branch:

Patch by Don Lipari, LLNL.
https://github.com/chaos/slurm/commit/4de11bf0a8cd18207a60e7d3e1fa7a6fde0da431

Move a log message before fork() call in the new spank prolog/epilog
logic. The original code could deadlock in glibc when getting the time
to include in the message.

If the select plugin does not allocate memory to a job, then do not
try to alloocate memory to the job step, without this the step allocation
will always fail