Message to controller will retry after SlurmctldTimeout period if
message to primary controller fails and backup controller returns
error indicating it is in backup mode.

nodes which are also in completing or down state.

Create the job hash table at slurmctld initialization.

WaitTime sets srun's default value for --wait. MaxJobCount sets
the maximum job count for slurmctld (replacing #define MAX_JOB_CNT).
MinJobAge sets the minimum job purrge age for slurmctld (replacing
#define MIN_JOB_AGE).

Major restructuring of the sinfo code.

with sinfo.

the node is DOWN. This gets cleaned up when the KILL_JOB RPC gets issued.

job, step, node, and partition info.
Add new function to free slurm_credential from RPC response.

job, step, node, and partition info.

noheader option, even if no jobs/steps.

SLURM_SORT, SLURM_STATES, and SLURM_USERS.

 o session manager waits for processes in its session, not just children.
   This fixes a problem when tasks are being attached to by a debugger.
 o Add elanhosts.[ch] for Elan host config file parsing. Used in slurm
   to install ElanID/hostname pairs into the kernel. Note that the format
   of the elanhosts config file has changed.

but don't wait for any reply. We remove the DOWN node from the job's
bitmap. As soon as the other nodes complete the KILL_JOB RPC, the job
transistions to some COMPLETED state.

If a node is down and not responding, don't bother to send a KILL_JOB
RPC to it. If that is the only node associated with a job, don't have
that job go through a COMPLETING state. It goes directly to a COMPLETED
state. Also preserve the NO_RESPOND flag associated with a node if its
state is changed via user request (e.g. scontrol).