based upon NodeAddr, ControllerAddr and BackupAddr configuration 
    parameters. Patch from Matthieu Hautreux, CEA.

    for better fault tolerance (e.g. killed and restarteed SLURM daemons on 
    compute node).

    nodes are configured in multiple leaf switches.

    launch tasks).

    with the --no-kill option.

  where needed for backwards compatibility

    file via RPC if topology/tree plugin is configured).

    which jobs can be submitted to this partition. From Matthieu Hautreux, CEA.

This patch abstracts the establishment of slurm user credentials
in slurmctld when SlurmUser != root. It adds a call to setgid()
to adopt the SLURM user's primary group, instead of running with
whatever primary group was in effect when invoking the slurmctld
executable. Also, errors in establishing user credentials are
fatal errors, instead of just warnings.

This is a minor security enhancement.

    --cpus-per-task option not properly handled.

We actually want to preserve the set_oomadj() function in slurm v1.4, but
have removed it completely from v1.3.

  the values of SLURMD_OOM_ADJ and SLURMSTEPD_OOM_ADJ environment
  variables. This can be used to prevent daemons being killed when
  a node's memory is exhausted.