def save_assigns_simu(self, jobs, resource_set):
print("save_assigns_simu")
for jid, job in iteritems(jobs):
jres_set = job.res_set
print("job.res_set before", jid, job.res_set)
r_ids = [resource_set.rid_o2i[roid] for roid in itvs2ids(jres_set)]
job.res_set = unordered_ids2itvs(r_ids)
self.assigned_jobs = jobs
def remove_gantt_resource_job(moldable_id, job_res_set, resource_set):
riods = itvs2ids(job_res_set)
resource_ids = [resource_set.rid_o2i[rid] for rid in riods]
db.query(GanttJobsResource)\
.filter(GanttJobsResource.moldable_id == moldable_id)\
.filter(~GanttJobsResource.resource_id.in_(tuple(resource_ids)))\
.delete(synchronize_session=False)
db.commit()
def save_assigns_redis_0(jobs, resource_set):
if len(jobs) > 0:
r = redis.Redis()
mld_id_start_time_s = []
for j in itervalues(jobs):
mld_id_start_time_s.append({"moldable_job_id": j.moldable_id, "start_time": j.start_time})
riods = itvs2ids(j.res_set)
str_mld_id_rids = ",".join(map(lambda x: str(resource_set.rid_o2i[x]), riods))
r.set(str(j.moldable_id), str_mld_id_rids)
db.session.execute(GanttJobsPrediction.__table__.insert(), mld_id_start_time_s)
def save_assigns_bulk_0(jobs, resource_set):
if len(jobs) > 0:
mld_id_start_time_s = []
mld_id_rid_s = []
for j in itervalues(jobs):
mld_id_start_time_s.append((j.moldable_id, j.start_time))
riods = itvs2ids(j.res_set)
mld_id_rid_s.extend(
[(j.moldable_id, resource_set.rid_o2i[rid]) for rid in riods])
with db.engine.connect() as to_conn:
cursor = to_conn.connection.cursor()
pg_bulk_insert(cursor, db['gantt_jobs_predictions'], mld_id_start_time_s,
('moldable_job_id', 'start_time'), binary=True)
pg_bulk_insert(cursor, db['queues'], mld_id_rid_s,
('moldable_job_id', 'resource_id'), binary=True)
def save_assigns_simu_and_default(self, jobs, resource_set):
print("save_assigns_simu_and_default........................")
# assigned_jobs = {}
for jid, job in iteritems(jobs):
sid = self.db_jid2s_jid[jid]
jobsimu = self.jobs[sid]
jres_set = job.res_set
r_ids = [resource_set.rid_o2i[roid] for roid in itvs2ids(jres_set)]
jobsimu.res_set = unordered_ids2itvs(r_ids)
print("save assign jid, sid, res_set: ", jid, " ", sid, " ", jobsimu.res_set)
jobsimu.start_time = job.start_time
jobsimu.walltime = job.walltime
# assigned_jobs[sid] = jobsimu
# self.assigned_jobs = assigned_jobs
return save_assigns(jobs, resource_set)
def save_assigns_redis_pipeline_0(jobs, resource_set):
print('# save_assigns_redis_pipeline_0')
t = 0
if len(jobs) > 0:
r = redis.Redis()
pipe = r.pipeline()
mld_id_start_time_s = []
for j in itervalues(jobs):
t0 = time.time()
mld_id_start_time_s.append(
{'moldable_job_id': j.moldable_id, 'start_time': j.start_time})
riods = itvs2ids(j.res_set)
str_mld_id_rids = ','.join(map(lambda x: str(resource_set.rid_o2i[x]), riods))
t += (time.time() - t0)
pipe.set(str(j.moldable_id), str_mld_id_rids)
db.session.execute(
GanttJobsPrediction.__table__.insert(), mld_id_start_time_s)
pipe.execute()
print("Cumlated mld_id_start_time_s.append time:", t)
def save_assigns_bulk(jobs, resource_set):
if len(jobs) > 0:
logger.debug("nb job to save: " + str(len(jobs)))
mld_id_start_time_s = []
mld_id_rid_s = []
for j in itervalues(jobs):
logger.debug("first job_id to save: " + str(j.id))
mld_id_start_time_s.append((j.moldable_id, j.start_time))
riods = itvs2ids(j.res_set)
mld_id_rid_s.extend(
[(j.moldable_id, resource_set.rid_o2i[rid]) for rid in riods])
logger.info("save assignements")
with db.engine.connect() as to_conn:
cursor = to_conn.connection.cursor()
pg_bulk_insert(cursor, db['gantt_jobs_predictions'], mld_id_start_time_s,
('moldable_job_id', 'start_time'), binary=True)
pg_bulk_insert(cursor, db['queues'], mld_id_rid_s,
('moldable_job_id', 'resource_id'), binary=True)
def save_assigns(jobs, resource_set):
# http://docs.sqlalchemy.org/en/rel_0_9/core/dml.html#sqlalchemy.sql.expression.Insert.values
if len(jobs) > 0:
logger.debug("nb job to save: " + str(len(jobs)))
mld_id_start_time_s = []
mld_id_rid_s = []
for j in itervalues(jobs):
logger.debug("first job_id to save: " + str(j.id))
mld_id_start_time_s.append(
{'moldable_job_id': j.moldable_id, 'start_time': j.start_time})
riods = itvs2ids(j.res_set)
mld_id_rid_s.extend(
[{'moldable_job_id': j.moldable_id,
'resource_id': resource_set.rid_o2i[rid]} for rid in riods])
logger.info("save assignements")
db.session.execute(
GanttJobsPrediction.__table__.insert(), mld_id_start_time_s)
db.session.execute(GanttJobsResource.__table__.insert(), mld_id_rid_s)
db.commit()
def gantt_init_with_running_jobs(plt, initial_time_sec, job_security_time):
#
# Determine Global Resource Intervals and Initial Slot
#
resource_set = plt.resource_set()
initial_slot_set = SlotSet((resource_set.roid_itvs, initial_time_sec))
logger.debug("Processing of processing of already handled reservations")
accepted_ar_jids, accepted_ar_jobs = \
get_waiting_reservations_already_scheduled(resource_set, job_security_time)
gantt_flush_tables(accepted_ar_jids)
# TODO Can we remove this step, below ???
# why don't use: assigned_resources and job start_time ??? in get_scheduled_jobs ???
logger.debug("Processing of current jobs")
current_jobs = get_jobs_in_multiple_states(['Running', 'toLaunch', 'Launching',
'Finishing', 'Suspended', 'Resuming'],
resource_set)
plt.save_assigns(current_jobs, resource_set)
#
# Resource availabilty (Available_upto field) is integrated through pseudo job
#
pseudo_jobs = []
for t_avail_upto in sorted(resource_set.available_upto.keys()):
itvs = resource_set.available_upto[t_avail_upto]
j = JobPseudo()
j.start_time = t_avail_upto
j.walltime = MAX_TIME - t_avail_upto
j.res_set = itvs
j.ts = False
j.ph = NO_PLACEHOLDER
pseudo_jobs.append(j)
if pseudo_jobs != []:
initial_slot_set.split_slots_jobs(pseudo_jobs)
#
# Get already scheduled jobs advanced reservations and jobs from more higher priority queues
#
# TODO?: Remove resources of the type specified in
# SCHEDULER_AVAILABLE_SUSPENDED_RESOURCE_TYPE
scheduled_jobs = plt.get_scheduled_jobs(
resource_set, job_security_time, initial_time_sec)
# retrieve ressources used by besteffort jobs
besteffort_rid2job = {}
for job in scheduled_jobs:
# print("job.id:", job.id, job.queue_name, job.types, job.res_set, job.start_time)
if 'besteffort' in job.types:
for r_id in itvs2ids(job.res_set):
besteffort_rid2job[r_id] = job
# Create and fill gantt
all_slot_sets = {'default': initial_slot_set}
if scheduled_jobs != []:
filter_besteffort = True
set_slots_with_prev_scheduled_jobs(all_slot_sets, scheduled_jobs,
job_security_time, initial_time_sec,
filter_besteffort)
return (all_slot_sets, scheduled_jobs, besteffort_rid2job)
def schedule_fifo_cycle(plt, queue="default", hierarchy_use=False):
assigned_jobs = {}
now = plt.get_time()
logger.info("Begin scheduling....now: " + str(now) + ", queue: " + queue)
#
# Retrieve waiting jobs
#
waiting_jobs, waiting_jids, nb_waiting_jobs = plt.get_waiting_jobs(queue)
if nb_waiting_jobs > 0:
logger.info("nb_waiting_jobs:" + str(nb_waiting_jobs))
for jid in waiting_jids:
logger.debug("waiting_jid: " + str(jid))
#
# Determine Global Resource Intervals
#
resource_set = plt.resource_set()
res_itvs = deepcopy(resource_set.roid_itvs)
#
# Get additional waiting jobs' data
#
job_security_time = int(config["SCHEDULER_JOB_SECURITY_TIME"])
plt.get_data_jobs(waiting_jobs, waiting_jids, resource_set, job_security_time)
#
# Remove resources used by running job
#
for job in plt.get_scheduled_jobs(resource_set, job_security_time, now):
if job.state == "Running":
res_itvs = sub_intervals(res_itvs, job.res_itvs)
#
# Assign resource to jobs
#
for jid in waiting_jids:
job = waiting_jobs[jid]
# We consider only one instance of resources request (no support for moldable)
(mld_id, walltime, hy_res_rqts) = job.mld_res_rqts[0]
if hierarchy_use:
# Assign resources which hierarchy support (uncomment)
itvs = find_resource_hierarchies_job(res_itvs, hy_res_rqts, resource_set.hierarchy)
else:
# OR assign resource by considering only resource_id (no hierarchy)
# and only one type of resource
(hy_level_nbs, constraints) = hy_res_rqts[0]
(h_name, nb_asked_res) = hy_level_nbs[0]
itvs_avail = intersec(constraints, res_itvs)
ids_avail = itvs2ids(itvs_avail)
if len(ids_avail) < nb_asked_res:
itvs = []
else:
itvs = unordered_ids2itvs(ids_avail[:nb_asked_res])
if (itvs != []):
job.moldable_id = mld_id
job.res_set = itvs
assigned_jobs[job.id] = job
res_itvs = sub_intervals(res_itvs, itvs)
else:
logger.debug("Not enough available resources, it's a FIFO scheduler, we stop here.")
break
#
# Save assignement
#
logger.info("save assignement")
plt.save_assigns(assigned_jobs, resource_set)
else:
logger.info("no waiting jobs")
def test_itvs2ids():
y = [(1, 1), (3, 5), (7, 7), (10, 12), (23, 23)]
r = [1, 3, 4, 5, 7, 10, 11, 12, 23]
a = itvs2ids(y)
assert a == r
