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.

Patch by Bill Brophy, Bull.

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

CRAY: Fix support for configuration with SlurmdTimeout=0 (never mark
node that is DOWN in ALPS as DOWN in SLURM).

Improve support for overlapping advanced reservations.
Patch from Bill Brophy, Bull.

already pinged it on startup the unresponding flag would be removed from
the frontend node.

mark front end node down.

Cray - For srun wrapper when creating a job allocation, set the default job
name to the executable file's name. Ignore leading directory names in the path.

permit the srun and salloc commands to be executed in the background
on Cray systems

Add new job state reason of "FrontEndDown" which applies only to Cray and
IBM BlueGene systems.

the queue when trying to place a larger than midplane job.

In cray/srun wrapper, only include aprun "-q" option when srun "--quiet"
option is used.

Patch from Aleksej Saushev.

is a convenience function in BSD and glibc that internally calls
the equivalent of

    int masterfd = open("/dev/ptmx", flags);
    grantpt (masterfd);
    unlockpt (masterfd);
    int slavefd = open (slave, O_RDRW|O_NOCTTY);

(in psuedocode)

On Linux, with some combinations of glibc/kernel (in this
case glibc-2.14/Linux-3.1), the equivalent of grantpt(3) was failing
in slurmstepd with EPERM, because the allocated pty was getting
root ownership instead of the user running the slurm job.

From the POSIX description of grantpt:

 "The grantpt() function shall change the mode and ownership of the
  slave pseudo-terminal device... The user ID of the slave shall
  be set to the real UID of the calling process..."

 http://pubs.opengroup.org/onlinepubs/007904875/functions/grantpt.html

This means that for POSIX-compliance, the real user id of slurmstepd
must be the user executing the SLURM job at the time openpty(3) is
called. Unfortunately, the real user id of slurmstepd at this
point is still root, and only the effective uid is set to the user.

This patch is a work-around that uses the (non-portable) setresuid(2)
system call to reset the real and effective uids of the slurmstepd
process to the job user, but keep the saved uid of root. Then after
the openpty(3) call, the previous credentials are reestablished
using the same call.

Fix for srun allocating running within existing allocation with --exclude
option and --nnodes count small enough to remove more nodes.

    > salloc -N 8
    salloc: Granted job allocation 1000008
    > srun -N 2 -n 2 --exclude=tux3 hostname
    srun: error: Unable to create job step: Requested node configuration is not available

Patch from Phil Eckert, LLNL.

Fix bug in step task distribution when nodes are not configured in numeric
order. Patch from Hongjia Cao, NUDT.

Fix bug when requeued batch job is scheduled to run on a different node
zero, but attemts job launch on old node zero causing fatal error
"Invalid host_index -1 for job #"

Avoid slurmctld abort due to bad pointer when setting an advanced
reservation MAINT flag if it contains no nodes (only licenses).

blocks are in an error state.

to give a correct priority on the first decay cycle after a restart of the
slurmctld. Patch from Martin Perry, Bull.

Lucero Palau.

This patch was previously applied to SLURM v2.4 and is being back-ported
due to problems being reported in SLURM v2.3. Original commit is here
https://github.com/SchedMD/slurm/commit/4c0eea7b8c20ccb1cacad51838a1ea8257cc637d

Set DEFAULT flag in partition structure when slurmctld reads the
configuration file. Patch from Rémi Palancher. Note the flag is set
when the information is sent via RPC for sinfo.