Bug 7445

Signed-off-by: Dominik Bartkiewicz <bart@schedmd.com>

When --batch=<feature> is used, the batch_host isn't chosen until the
job is being launched -- because the features could be different on boot
(e.g. KNL nodes). Thus if the job is allocated nodes that need to be
booted, it needs to wait till they are all booted so it can make a
decision at launch time.

Bug 7445

Signed-off-by: Dominik Bartkiewicz <bart@schedmd.com>

This is a continuation to 7da439b4

Bug 7445

The get_default_acct was truncating the account name for long account
names. This commit uses -P/--parsable2 to avoid it.

Bug 7369.

Bug 7467.

Only do so once the task actually finishes. Otherwise, a requeued task
could set an incorrect max_exit_code even if completed with exit code 0
after re-running again, leading to problems with i.e. other jobs with an
afterok type of dependency on such array relying on the incorrectly set
max_exit_code.

Bug 7552.

We only remove from registered_clusters if conn->rem_port != 0,
so only add to it if the same.

Bug 5213

Signed-off-by: Alejandro Sanchez <alex@schedmd.com>

The previous log message implied that you should never use the topology
plugin where no switch could reach all nodes through its descendants.
However, this is a valid configuration where sites may not want jobs
spanning across certain switches, so we've softened the language in the
log message.

Bug 7466.

In case of features like cpu&fastio&[knl|westmere] additional bit_or
resulted in returning something like (cpu&fastio)|knl|westmere, which
is obviously wrong. XOR/XAND features are handled properly in
_get_req_features.

Bug 7378

Display nodenames instead of bitmap ranges

We changed FAQ in 4cea931c we replaced stop/start of slurmd with just
restart, but the example now suggest to use systemctl start which will
actually do nothing in case of started slurmd.

Handle situation where a slurmctld tries to communicate with slurmdbd more than once at the same time.

What can happen here is the slurmdbd/slurmctld connection gets hung up
somehow.  If the slurmctld is restarted a new connection is made along
side the old connection.  When the old connection gets unwedged the old
connection will clear out the registration of the slurmctld making it so
no updates are sent to that slurmctld.

What this does is checks for old connections when a registration message
comes in.  If we find one we print error set the rem_port = 0 and
remove it from the list.  This makes it so when it gets unwedged we just
close the socket instead of remove the registration.

Bug 5213

Bug 7360.

in track scripts code.

Bug 7360


Signed-off-by: Alejandro Sanchez <alex@schedmd.com>

The implementation of priority_p_job_end in priority/multifactor
expects the job state to be set to complete or completing in order to
properly remove some job usage from the assoc and qos. This must be
simulated by the pack job run check code, or the check-time usage is not
removed.

Bug 7284

Bug 7428

Use of scontrol wait_job in slurmctld will result in prolog
hanging since the command will complete only when PrologSlurmctld
is completed. It's a deadlock.

Bug 7428.

It wasn't properly set under certain conditions.

Bug 7276

"ANY" is the canonical and most accurate value identifier for
PARTITION_ENFORCE_ANY although "Yes", "Up", "True" and "1" continue
being parsed and accepted as equivalent values for retro-compatibility
purposes with the initial commit edf3880c.

Bug 7248.

Bug 7410.

Continuation of 884c0191.

Bug 7362.

Bug 6769

Consider jobs in COMPLETING state as being available immediatley for
a job will-run evaluation. This assumes the completion will happen
very soon after the test is run.

bug 6769

All of the select plugins were performing a duplicate resource free
for jobs in completing state when performing a will-run test for
new jobs. This would frequently result in underflow messages.

Bug 6769

Building off the prevoius commit, spank_option_getopt() is now valid in
more functions than before, so we document and enforce from where
spank_option_getopt() can safely be called and return ESPANK_NOT_AVAIL
if it is called from any invalid SPANK context.

Bug 7065.

When spank option callbacks are called, the options are added to a cache
in memory so that spank_option_getopt() can retrieve the options when
called. However, this was only happening when callbacks were called from
the local context, so we make sure that the options are added to the
cache when the callbacks are called from the remote context as well.

Bug 7065.

job_test_resv() - return ESLURM_RESERVATION_MAINT if all nodes
from the job partition are in a maintenance reservation.

Bug 7362.

If job requests feature(s) but the required nodes are in an advanced
reservation.

Bug 7362.

Previously the UnavailableNodes list could include nodes that don't
belong to the job partition.

Bug 7362.