with "Shared=yes" configuration failed to function as desired. (gnats:459).

with FastSchedule==0 configuration.

    calls, improves performance and eliminates realloc failure on
    RH EL3 under extremely heavy workload apparently due to memory
    fragmentation

bit_alloc calls.

time or node limits.

  (which exhaust available memory).
Restructure slurmctld agent logic for higher throughput.

(gnats:381)

them when partition limits change.

  clusters with FastSchedule configured off
* Only return DOWN nodes to service if the reason for them being in
  that state is non-responsiveness and ReturnToService configured on
* Some general code clean-up

sharing via node record of job count (0 | 1) and bitmap of nodes which
permit sharing. Previous logic could permit a job accepting shared nodes
to be scheduled on a node that already had a running job not accepting
shared nodes.

gracefully kill all jobs allocated resources on those nodes,
gracefully kill all pending jobs that require those nodes,
leave pending jobs that exclude those nodes but ignore those nodes.
Added "best_effort" argument to node_name2bitmap() function.
Fix potential memory leak when maui scheduler interface resets
the required nodes.
(gnats:342)

attempt failure (i.e. convert to event driven instead of just polling).
It does still have a 2-second timer to keep the overhead reasonable.

be changed whenever a node responded to a "ping" and other insignificant
events, which resulted in the backfill scheduling running more frequently
than required.

purge the request and job if/when the node changes to state DOWN.

using node hostlist format.

Job's priority would be reset if it could not be immediately scheduled
AND if the best-fit node selection picked a DOWN node.

lost of an EPILOG_COMPLETE message.

Fatal() no longer calls abort(), but terminates job using exit(1).

This is a problem on PVC, where nodes are regularly DRAINED.

transition to DRAINED state.

right away (before running scheduling function).

nodes which are not available (DOWN or DRAIN). This will prevent them from
blocking other jobs from using the nodes which are available (i.e. over-ride
FIFO scheduling).
(gnats:279)

resources to satisfy the request).

batch_job_launch RPC, then deallocate those resources and requeue the job.
If a node registers and fails to show a batch job that should have a
script running there (node zero of allocation), then consider the job
complete.

Perform some general code clean-up in those areas.

shutdown only).

while fail/if too many then fatal/sleep and retry

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.