def test_is_valid(self):
"""
There are some things that are required. Check these things here!
"""
job = ModelJob()
self.assertFalse(job.is_valid())
# If all of the required arguments are supplied, this should result in a valid job
ts_complete_set = {
tsk: TimeSignal.from_values(tsk, [0., 0.1], [1., 999.])
for tsk in time_signal_names
}
valid_args = {
'time_start': 0,
'duration': 0.2,
'ncpus': 1,
'nnodes': 1,
'timesignals': ts_complete_set
}
self.assertTrue(ModelJob(**valid_args).is_valid())
# If any of the supplied arguments are missing, this should invalidate things
for k in valid_args.keys():
invalid_args = valid_args.copy()
del invalid_args[k]
self.assertTrue(ModelJob(**valid_args).is_valid())
def test_merge(self):
"""
Can we merge multiple ModelJobs.
n.b. Currently this only supports time signals!
TODO: Merge the non-time-signal data.
"""
# n.b. non-zero values. Zero time signals are ignored.
kb_read1 = TimeSignal.from_values('kb_read', [0.0], [1.0], priority=8)
kb_read2 = TimeSignal.from_values('kb_read', [0.0], [1.0], priority=10)
kb_write1 = TimeSignal.from_values('kb_write', [0.0], [1.0],
priority=8)
# Test that we take the union of the available time series
job1 = ModelJob(label="label1", timesignals={'kb_read': kb_read1})
job2 = ModelJob(label="label1", timesignals={'kb_write': kb_write1})
job1.merge(job2)
self.assertEqual(len(job1.timesignals), len(signal_types))
self.assertEqual(job1.timesignals['kb_read'], kb_read1)
self.assertEqual(job1.timesignals['kb_write'], kb_write1)
# (The other time signals should still be None)
for ts_name in signal_types:
if ts_name in ['kb_read', 'kb_write']:
continue
self.assertIn(ts_name, job1.timesignals)
self.assertIsNone(job1.timesignals[ts_name])
# check that when merging we take the signal with highest priority index
job1 = ModelJob(label="label1", timesignals={'kb_read': kb_read1})
job2 = ModelJob(label="label1", timesignals={'kb_read': kb_read2})
job1.merge(job2)
self.assertEqual(job1.timesignals['kb_read'], kb_read2)
def test_merge_fails_different_label(self):
"""
We should not be able to merge two jobs with differing labels, as these don't correspond to the same overall job
"""
job1 = ModelJob(label="a-label-1")
job2 = ModelJob(label="a-label-2")
self.assertRaises(AssertionError, lambda: job1.merge(job2))
def model_job(self):
"""
Return a ModelJob from the supplied information
"""
assert len(self.tasks) > 0
# Combine the task metadata. We take the first observed task time as the start time, and the last observed
# time as the end time.
time_start = self.time_start
time_end = max([t.time_end for t in self.tasks])
duration = time_end - time_start
# The number of tasks and hosts may also vary, as there may be multiple aprun/mpirun/mpiexec calls profiled
ntasks = max([t.ntasks for t in self.tasks])
nhosts = max([t.nhosts for t in self.tasks])
# TODO: We want to capture multi-threading as well as multi-processing somewhere3
return ModelJob(
job_name=self.jobname,
user_name=self.user,
queue_name=self.queue_type,
time_queued=self.time_queued,
time_start=self.time_start,
duration=duration,
ncpus=ntasks,
nnodes=nhosts,
stdout=self.stdout,
label=self.label,
timesignals=self.model_time_series(),
)
def model_jobs(self):
for job in self.joblist:
assert isinstance(job, IngestedJob)
assert not job.timesignals
# yield ModelJob(
# time_start=job.time_start - self.global_start_time,
# duration=job.time_end-job.time_start,
# ncpus=job.ncpus,
# nnodes=job.nnodes,
# scheduler_timing=job.time_queued,
# stdout=job.stdout
# )
yield ModelJob(
job_name=job.jobname,
user_name=job.user,
cmd_str=job.cmd_str,
queue_name=job.queue_type,
time_queued=job.time_queued,
time_start=job.time_start,
duration=job.time_end - job.time_start,
ncpus=job.ncpus,
nnodes=job.nnodes,
scheduler_timing=job.time_queued,
stdout=job.stdout,
label=None,
)
def test_merge_ignores_empty_timesignals(self):
"""
When merging in time signals from another job, if there is no data in the "other" time signal, then it
should be ignored for merging purposes.
:return:
"""
kb_read = TimeSignal.from_values('kb_read', [0.0], [1.0])
kb_write = TimeSignal.from_values('kb_write', [0.0],
[0.0]) # n.b. zero data
job1 = ModelJob(label="label1", timesignals={'kb_read': kb_read})
job2 = ModelJob(label="label1", timesignals={'kb_write': kb_write})
self.assertIsNone(job1.timesignals['kb_write'])
self.assertIsNotNone(job2.timesignals['kb_write'])
job1.merge(job2)
self.assertIsNone(job1.timesignals['kb_write'])
def test_reject_mislabelled_time_signals(self):
"""
The initialisation routine should reject invalid time signals in a model job.
"""
self.assertRaises(
ModellingError, lambda: ModelJob(timesignals={
'kb_write':
TimeSignal.from_values('kb_read', [0.0], [0.0]),
}))
def test_reanimation_kprofile(self):
"""
The purpose of the KProfile is to be able to (re-)animate ModelJobs from the input data.
"""
valid = {
"version":
1,
"tag":
"KRONOS-KPROFILE-MAGIC",
"created":
"2016-12-14T09:57:35Z", # Timestamp in strict rfc3339 format.
"uid":
1234,
"workload_tag":
"A-tag",
"profiled_jobs": [{
"time_start": 537700,
"time_queued": 99,
"duration": 147,
"ncpus": 72,
"nnodes": 2,
"time_series": {
"kb_read": {
"times": [0.01, 0.02, 0.03, 0.04],
"values": [15, 16, 17, 18],
"priority": 10
}
}
}]
}
pf = ProfileFormat.from_file(StringIO(json.dumps(valid)))
jobs = [ModelJob.from_json(j) for j in pf.profiled_jobs]
self.assertEqual(len(jobs), 1)
self.assertIsInstance(jobs[0], ModelJob)
self.assertEqual(jobs[0].time_start, 537700)
self.assertEqual(jobs[0].time_queued, 99)
self.assertEqual(jobs[0].duration, 147)
self.assertEqual(jobs[0].ncpus, 72)
self.assertEqual(jobs[0].nnodes, 2)
self.assertEqual(len(jobs[0].timesignals), len(signal_types))
self.assertIn('kb_read', jobs[0].timesignals)
for name, signal in jobs[0].timesignals.items():
if name == 'kb_read':
self.assertIsInstance(signal, TimeSignal)
self.assertTrue(
all(x1 == x2 for x1, x2 in zip(signal.xvalues,
[0.01, 0.02, 0.03, 0.04])))
self.assertTrue(
all(y1 == y2
for y1, y2 in zip(signal.yvalues, [15, 16, 17, 18])))
else:
self.assertIsNone(signal)
def test_initialisation(self):
# Test some defaults
job = ModelJob()
for attr in ['time_start', 'ncpus', 'nnodes', 'duration', 'label']:
self.assertTrue(hasattr(job, attr))
self.assertIsNone(getattr(job, attr))
for ts_name in signal_types:
self.assertIn(ts_name, job.timesignals)
self.assertIsNone(job.timesignals[ts_name])
# Test that we can override specified fields
job = ModelJob(timesignals={
'kb_read':
TimeSignal.from_values('kb_read', [0.0], [0.0]),
'kb_write':
TimeSignal.from_values('kb_write', [0.0], [0.0]),
},
time_start=123,
ncpus=4,
nnodes=5,
duration=678,
label="a-label")
self.assertEqual(job.time_start, 123)
self.assertEqual(job.ncpus, 4)
self.assertEqual(job.nnodes, 5)
self.assertEqual(job.duration, 678)
self.assertEqual(job.label, "a-label")
self.assertIsInstance(job.timesignals['kb_read'], TimeSignal)
self.assertIsInstance(job.timesignals['kb_write'], TimeSignal)
for ts_name in signal_types:
if ts_name in ['kb_read', 'kb_write']:
continue
self.assertIn(ts_name, job.timesignals)
self.assertIsNone(job.timesignals[ts_name])
# Test that we cannot override non-specific fields
self.assertRaises(ModellingError, lambda: ModelJob(invalid=123))
def model_job(self):
"""
Return a model job from this job
:return:
"""
return ModelJob(time_start=0,
duration=self.max_duration(),
ncpus=self.n_procs,
nnodes=self.n_nodes,
timesignals=self.timesignals,
label=self.name)
def model_jobs(self):
for job in self.joblist:
assert isinstance(job, IngestedJob)
yield ModelJob(
job_name=job.jobname,
user_name=job.user,
queue_name=job.queue_type,
cmd_str=job.cmd_str,
time_queued=job.time_queued,
time_start=job.time_start,
duration=job.time_end - job.time_start,
ncpus=job.ncpus,
nnodes=job.nnodes,
stdout=job.stdout,
label=self.json_label_map[os.path.basename(job.jobname)]
if self.json_label_map else None,
timesignals=job.timesignals,
)
def model_job(self):
"""
Return a ModelJob from the supplied information
"""
if float(self.log_version) <= 2.0:
raise DarshanLogReaderError("Darshan log version unsupported")
return ModelJob(
job_name=self.filename,
user_name=None, # not provided
cmd_str=self.exe_cmd,
queue_name=None, # not provided
time_queued=None, # not provided
time_start=self.time_start,
duration=self.time_end - self.time_start,
ncpus=self.nprocs,
nnodes=None, # not provided
stdout=None, # not provided
label=self.label,
timesignals=self.model_time_series(),
)
def model_jobs(self):
for job in self.joblist:
# assert isinstance(job, IngestedJob)
assert not job.timesignals
# if job.time_created >= 0:
# submit_time = job.time_created - self.global_created_time
# else:
# submit_time = job.time_start - self.global_start_time
yield ModelJob(
job_name=job.jobname,
user_name=job.user,
cmd_str=job.cmd_str,
queue_name=job.queue_type,
time_queued=job.time_queued,
time_start=job.time_start,
duration=job.time_end - job.time_start,
ncpus=job.ncpus,
nnodes=job.nnodes,
stdout=job.stdout,
label=None,
)
def test_workload_fillin_default(self):
"""
Test the correct assignment of user-defined time-series
:return:
"""
io_metrics = ['kb_read', 'kb_write', 'n_read', 'n_write']
# create 2 random jobs (with ONLY io metrics)
valid_args_1 = {
'time_start': 0.1,
'duration': 0.2,
'ncpus': 1,
'nnodes': 1,
'timesignals': {
tsk: TimeSignal.from_values(tsk, np.random.rand(10),
np.random.rand(10))
for tsk in io_metrics
}
}
job1 = ModelJob(**valid_args_1)
valid_args_2 = {
'time_start': 0.1,
'duration': 0.2,
'ncpus': 1,
'nnodes': 1,
'timesignals': {
tsk: TimeSignal.from_values(tsk, np.random.rand(10),
np.random.rand(10))
for tsk in io_metrics
}
}
job2 = ModelJob(**valid_args_2)
test_workload = Workload(jobs=[job1, job2], tag='wl_2jobs')
# ---------------------- fill in config -----------------------
filling_funct_config = [{
"type": "step",
"name": "step-1",
"x_step": 0.5
}, {
"type":
"custom",
"name":
"custom-1",
"x_values": [0, 0.1, 0.15, 0.3333, 0.5, 0.8, 0.9, 1.0],
"y_values": [0, 0.1, 0.2, 0.3, 0.5, 0.8, 0.9, 1.0]
}]
# Values to assign to all the unspecified metrics
default_config = {
"type": "fill_missing_entries",
"apply_to": ["wl_2jobs"],
"priority": 0,
"metrics": {
"kb_collective": [100, 101],
"n_collective": [100, 101],
"kb_pairwise": {
"function": "step-1",
"scaling": 1000.0
},
"n_pairwise": {
"function": "custom-1",
"scaling": 1000.0
},
"flops": [100, 101],
}
}
# update the filling config with the user-defined functions
default_config.update({"user_functions": filling_funct_config})
# Apply the user defaults to the workloads
workloads = [test_workload]
filler = StrategyUserDefaults(workloads)
filler.apply(default_config)
# test that the IO metrics are within the random range used [0,1]
for j in workloads[0].jobs:
self.assertTrue(
all([0.0 < x < 1.0 for x in j.timesignals['n_write'].xvalues]))
self.assertTrue(
all([0.0 < x < 1.0 for x in j.timesignals['n_write'].yvalues]))
self.assertTrue(
all([0.0 < x < 1.0
for x in j.timesignals['kb_write'].xvalues]))
self.assertTrue(
all([0.0 < x < 1.0
for x in j.timesignals['kb_write'].yvalues]))
self.assertTrue(
all([0.0 < x < 1.0 for x in j.timesignals['n_read'].xvalues]))
self.assertTrue(
all([0.0 < x < 1.0 for x in j.timesignals['n_read'].yvalues]))
self.assertTrue(
all([0.0 < x < 1.0 for x in j.timesignals['kb_read'].xvalues]))
self.assertTrue(
all([0.0 < x < 1.0 for x in j.timesignals['kb_read'].yvalues]))
# test that the user-defined metrics are within the random range chosen [0,1]
for j in workloads[0].jobs:
self.assertTrue(
all([100 < x < 101 for x in j.timesignals['flops'].yvalues]))
self.assertTrue(
all([
100 < x < 101
for x in j.timesignals['n_collective'].yvalues
]))
self.assertTrue(
all([
100 < x < 101
for x in j.timesignals['kb_collective'].yvalues
]))
# test that the user-defined functions are being applied as expected
for j in workloads[0].jobs:
# values vs expected
val_exp = zip(j.timesignals['n_pairwise'].yvalues,
[0, 0.1, 0.2, 0.3, 0.5, 0.8, 0.9, 1.0])
self.assertTrue(all([x == y * 1000. for x, y in val_exp]))
# and the step function
self.assertTrue(
all([(x == 0 or x == 1000.)
for x in j.timesignals['kb_pairwise'].yvalues]))
def test_generator(self):
"""
The configuration object should have some sane defaults
"""
# If all of the required arguments are supplied, this should result in a valid job
ts_complete_set = {
tsk: TimeSignal.from_values(tsk, [0., 0.1], [1., 999.])
for tsk in time_signal_names
}
valid_args = {
'time_start': 0.1,
'duration': 0.2,
'ncpus': 1,
'nnodes': 1,
'timesignals': ts_complete_set
}
ts_complete_set_2 = {
tsk: TimeSignal.from_values(tsk, [0., 0.1], [1., 444.])
for tsk in time_signal_names
}
valid_args_2 = {
'time_start': 0.1,
'duration': 0.2,
'ncpus': 1,
'nnodes': 1,
'timesignals': ts_complete_set_2
}
# check that it is a valid job
job1 = ModelJob(**valid_args)
job1.label = "job1"
job2 = ModelJob(**valid_args_2)
job2.label = "job2"
job3 = ModelJob(**valid_args)
job3.label = "job3"
job4 = ModelJob(**valid_args_2)
job4.label = "job4"
job5 = ModelJob(**valid_args)
job5.label = "job5"
input_jobs = [job1, job2, job3, job4, job5]
# diversify the time start..
for jj, job in enumerate(input_jobs):
job.time_start += jj * 0.1
for job in input_jobs:
self.assertTrue(job.is_valid())
config_generator = {
"type": "cluster_and_spawn",
"job_clustering": {
"type": "Kmeans",
"rseed": 0,
"apply_to": ["test_wl_0"],
"ok_if_low_rank": True,
"max_iter": 100,
"max_num_clusters": 3,
"delta_num_clusters": 1,
"num_timesignal_bins": 1,
"user_does_not_check": True
},
"job_submission_strategy": {
"type": "match_job_pdf_exact",
"n_bins_for_pdf": 20,
"submit_rate_factor": 8,
"total_submit_interval": 60,
"random_seed": 0
}
}
# select the appropriate workload_filling strategy
workloads = [
Workload(jobs=input_jobs, tag='test_wl_0'),
Workload(jobs=input_jobs, tag='test_wl_1'),
Workload(jobs=input_jobs, tag='test_wl_2')
]
workload_modeller = workload_modelling_types[config_generator["type"]](
workloads)
workload_modeller.apply(config_generator)
# get the newly created set of (modelled) workloads
workload_set = workload_modeller.get_workload_set()
# make sure that we are creating only one workload
self.assertEqual(len(workload_set.workloads), 1)
# ---- check that all the jobs are correctly formed.. ----
# check that each job has time-signals as expected..
for job in workload_set.workloads[0].jobs:
self.assertTrue(hasattr(job, "timesignals"))
# check that each job has all the time-signals at this point..
for job in workload_set.workloads[0].jobs:
self.assertTrue(
all([k in job.timesignals.keys() for k in time_signal_names]))
def generate_jobs(self):
logger.info("====> Generating jobs from sub-workload: {}, "
"that has {} jobs".format(self.wl_clusters['source-workload'],
len(self.wl_clusters['jobs_for_clustering'])))
start_times_vec_sa, _, _ = self.schedule_strategy.create_schedule()
np.random.seed(self.config["job_submission_strategy"]['random_seed'])
n_modelled_jobs = len(start_times_vec_sa)
n_clusters = self.wl_clusters['cluster_matrix'].shape[0]
clustered_jobs_all = self.wl_clusters['jobs_for_clustering']
clustered_jobs_labels = self.wl_clusters['labels']
# this is the fraction of jobs that need to be taken from eah cluster..
n_job_cluster = len(self.wl_clusters['jobs_for_clustering'])
model_job_fraction = min(1.0, n_modelled_jobs/float(n_job_cluster))
chosen_model_jobs = []
vec_clust_indexes = np.asarray([], dtype=int)
for ic in range(n_clusters):
# jobs in this cluster
jobs_in_cluster = np.asarray(clustered_jobs_all)[clustered_jobs_labels == ic]
jobs_in_cluster_idxs = np.arange(len(jobs_in_cluster))
n_jobs_to_generate_from_cluster = max(1, int(len(jobs_in_cluster)*model_job_fraction))
if n_jobs_to_generate_from_cluster > len(jobs_in_cluster):
n_jobs_to_generate_from_cluster = len(jobs_in_cluster)
# take a random sample of jobs in cluster (according to job fraction)
jobs_in_cluster_sampled_idx = np.random.choice(jobs_in_cluster_idxs,
n_jobs_to_generate_from_cluster,
replace=False)
jobs_in_cluster_for_generation = []
for idxj in jobs_in_cluster_sampled_idx:
jobs_in_cluster_for_generation.append(jobs_in_cluster[idxj])
# chosen_model_jobs.extend(jobs_in_cluster[jobs_in_cluster_sampled_idx])
chosen_model_jobs.extend(jobs_in_cluster_for_generation)
vec_clust_indexes = np.append(vec_clust_indexes,
np.ones(n_jobs_to_generate_from_cluster, dtype=int)*ic)
# generates model jobs as needed
generated_model_jobs = []
for cc, job in enumerate(chosen_model_jobs):
job_copy = copy.deepcopy(job)
# assign this job a start time (if more jobs are created,
# the start time is chosen randomly within the start times..)
if cc < len(start_times_vec_sa):
start_time = start_times_vec_sa[cc]
else:
start_time = start_times_vec_sa[np.random.randint(0, len(start_times_vec_sa), 1)[0]]
# choose a number of cpu for this job
if not job.ncpus:
logger.warning("job ID {} had not ncpu specified, "
"set to default ncpu=1 instead".format(cc))
job_ncpus = 1
else:
job_ncpus = job.ncpus
# choose a number of nodes for this job
if not job.nnodes:
job_nnodes = 1
else:
job_nnodes = job.nnodes
# Spawn and append the model job
job = ModelJob(
time_start=start_time,
duration=None,
ncpus=job_ncpus,
nnodes=job_nnodes,
timesignals=job_copy.timesignals,
label="job-{}".format(cc)
)
generated_model_jobs.append(job)
n_cluster_jobs = len(self.wl_clusters['jobs_for_clustering'])
n_job_ratio = len(generated_model_jobs)/float(n_cluster_jobs) * 100.
logger.info("<==== Generated {} jobs (#job ratio = {:.2f}%)".format(
len(generated_model_jobs), n_job_ratio))
return generated_model_jobs, vec_clust_indexes.tolist()
def generate_jobs(self):
logger.info("Generating jobs from cluster: {}, "
"that has {} jobs".format(
self.wl_clusters['source-workload'],
len(self.wl_clusters['jobs_for_clustering'])))
start_times_vec_sa, _, _ = self.schedule_strategy.create_schedule()
# Random vector of cluster indexes
n_modelled_jobs = len(start_times_vec_sa)
np.random.seed(self.config["job_submission_strategy"].get(
'random_seed', 0))
vec_clust_indexes = np.random.randint(
self.wl_clusters['cluster_matrix'].shape[0], size=n_modelled_jobs)
# Mean NCPU in cluster (considering jobs in cluster)
jobs_all = self.wl_clusters['jobs_for_clustering']
lab_all = np.asarray(self.wl_clusters['labels'])
# jobs in each cluster
jobs_in_each_cluster = {
cl: np.asarray(jobs_all)[lab_all == cl]
for cl in set(lab_all)
}
# mean #CPUS in each cluster (from jobs for which ncpus is available, otherwise 1)
mean_cpus = {
cl_id: np.mean([job.ncpus if job.ncpus else 1 for job in cl_jobs])
for cl_id, cl_jobs in jobs_in_each_cluster.items()
}
# mean #NODES in each cluster (from jobs for which nnodes is available, otherwise 1)
mean_nodes = {
cl_id:
np.mean([job.nnodes if job.nnodes else 1 for job in cl_jobs])
for cl_id, cl_jobs in jobs_in_each_cluster.items()
}
# loop over the clusters and generates jobs as needed
generated_model_jobs = []
for cc, cl_idx in enumerate(vec_clust_indexes):
ts_dict = {}
row = self.wl_clusters['cluster_matrix'][cl_idx, :]
ts_yvalues = np.split(row, len(time_signal_names))
for tt, ts_vv in enumerate(ts_yvalues):
ts_name = time_signal_names[tt]
ts = TimeSignal(ts_name).from_values(ts_name,
np.arange(len(ts_vv)),
ts_vv)
ts_dict[ts_name] = ts
job = ModelJob(time_start=start_times_vec_sa[cc],
job_name="job-{}-cl-{}".format(cc, cl_idx),
duration=None,
ncpus=mean_cpus[cl_idx],
nnodes=mean_nodes[cl_idx],
timesignals=ts_dict,
label="job-{}-cl-{}".format(cc, cl_idx))
generated_model_jobs.append(job)
n_sa = len(generated_model_jobs)
n_job_ratio = n_sa / float(len(
self.wl_clusters['jobs_for_clustering'])) * 100.
logger.info(
"====> Generated {} jobs from cluster (#job ratio = {:.2f}%)".
format(n_sa, n_job_ratio))
return generated_model_jobs, vec_clust_indexes
def test_workload_fillin_match(self):
"""
Test the metrics assignment through job name (label) matching
:return:
"""
# ------------ verify global time signals -------------------
valid_args_1 = {
'job_name': "blabla_weird_name",
'time_start': 0.1,
'duration': 0.222,
'ncpus': 1,
'nnodes': 1,
'timesignals': {
tsk: TimeSignal.from_values(tsk, np.random.rand(10),
np.arange(10) * 2)
for tsk in time_signal_names
}
}
job1 = ModelJob(**valid_args_1)
valid_args_2 = {
'job_name': "job_match",
'time_start': 0.1,
'duration': 0.333,
'ncpus': 1,
'nnodes': 1,
'timesignals': {}
}
job2 = ModelJob(**valid_args_2)
# ------ target workload (that will receive the time metrics..)
target_wl = Workload(jobs=[job1, job2], tag='target_workload')
# ---------- source workload
valid_args_3 = {
'job_name': "job_match",
'time_start': 0.1,
'duration': 0.333,
'ncpus': 1,
'nnodes': 1,
'timesignals': {
tsk: TimeSignal.from_values(tsk, np.random.rand(10),
np.random.rand(10))
for tsk in time_signal_names
}
}
job3 = ModelJob(**valid_args_3)
source_wl = Workload(jobs=[job3], tag='wl_match_source')
# filler config
filler_config = {
"type": "match_by_keyword",
"priority": 0,
"keywords": ["job_name"],
"similarity_threshold": 0.3,
"source_workloads": ["wl_match_source"],
"apply_to": ["target_workload"]
}
# Apply the user defaults to the workloads
workloads = [target_wl, source_wl]
filler = StrategyMatchKeyword(workloads)
filler.apply(filler_config)
# for ts_k, ts_v in job3.timesignals.iteritems():
# print "JOB3:{}:{}".format(ts_k, ts_v.yvalues)
#
# for ts_k, ts_v in target_wl.jobs[1].timesignals.iteritems():
# print "TRG_J1:{}:{}".format(ts_k, ts_v.yvalues)
self.assertTrue(
all([
all(ys == yt for ys, yt in zip(job3.timesignals[ts_k].yvalues,
ts_v.yvalues))
for ts_k, ts_v in target_wl.jobs[1].timesignals.items()
]))
def test_workload_data(self):
# If all of the required arguments are supplied, this should result in a valid job
ts_complete_set = {
tsk: TimeSignal.from_values(tsk, [0., 0.1], [1., 999.])
for tsk in time_signal_names
}
valid_args = {
'time_start': 0.1,
'duration': 0.2,
'ncpus': 1,
'nnodes': 1,
'timesignals': ts_complete_set
}
# check that it is a valid job
job1 = ModelJob(**valid_args)
job2 = ModelJob(**valid_args)
job3 = ModelJob(**valid_args)
job4 = ModelJob(**valid_args)
job5 = ModelJob(**valid_args)
input_jobs = [job1, job2, job3, job4, job5]
# diversify the time start..
for jj, job in enumerate(input_jobs):
job.time_start += jj * 0.1
for job in input_jobs:
self.assertTrue(job.is_valid())
# create a workload with 5 model jobs
test_workload = Workload(jobs=input_jobs, tag='test_wl')
# -- verify that all the jobs in workload are actually the initial jobs provided --
self.assertTrue(
all(job is input_jobs[jj]
for jj, job in enumerate(test_workload.jobs)))
# ------------ verify sums of timesignals -------------------
for ts_name in signal_types:
ts_sum = 0
for j in input_jobs:
ts_sum += sum(j.timesignals[ts_name].yvalues)
# verify the sums..
self.assertEqual(ts_sum,
test_workload.total_metrics_sum_dict[ts_name])
# ------------ verify global time signals -------------------
valid_args_1 = {
'time_start': 0.1,
'duration': 0.222,
'ncpus': 1,
'nnodes': 1,
'timesignals': {
tsk: TimeSignal.from_values(tsk, np.random.rand(10),
np.random.rand(10))
for tsk in time_signal_names
}
}
job1 = ModelJob(**valid_args_1)
valid_args_2 = {
'time_start': 0.1,
'duration': 0.333,
'ncpus': 1,
'nnodes': 1,
'timesignals': {
tsk: TimeSignal.from_values(tsk, np.random.rand(10),
np.random.rand(10))
for tsk in time_signal_names
}
}
job2 = ModelJob(**valid_args_2)
test_workload = Workload(jobs=[job1, job2], tag='wl_2jobs')
for job in [job1, job2]:
for ts in signal_types:
self.assertTrue(
all(v + job.time_start in
test_workload.total_metrics_timesignals[ts].xvalues
for v in job.timesignals[ts].xvalues))
self.assertTrue(
all(v in
test_workload.total_metrics_timesignals[ts].yvalues
for v in job.timesignals[ts].yvalues))
def test_merge_rejects_mislabelled_time_signals(self):
"""
Test that the merging routine checks the labelling validity. Both ways around.
:return:
"""
kb_read = TimeSignal.from_values('kb_read', [0.0], [1.0])
kb_write = TimeSignal.from_values('kb_read', [0.0],
[1.0]) # n.b. mislabelled
job1 = ModelJob(label="label1", timesignals={'kb_read': kb_read})
job2 = ModelJob(label="label1")
job2.timesignals['kb_write'] = kb_write
self.assertRaises(ModellingError, lambda: job1.merge(job2))
# And the other way around
job2 = ModelJob(label="label1", timesignals={'kb_read': kb_read})
job1 = ModelJob(label="label1")
job1.timesignals['kb_write'] = kb_write
self.assertRaises(ModellingError, lambda: job1.merge(job2))
def test_splitter(self):
# -------------- prepare a couple of dummy jobs ---------------
# If all of the required arguments are supplied, this should result in a valid job
ts_complete_set = {
tsk: TimeSignal.from_values(tsk, [0., 0.1], [1., 999.])
for tsk in time_signal_names
}
ts_complete_set_2 = {
tsk: TimeSignal.from_values(tsk, [0., 0.1], [1., 444.])
for tsk in time_signal_names
}
valid_args = {
'time_start': 0.1,
'duration': 0.2,
'ncpus': 1,
'nnodes': 1,
'timesignals': ts_complete_set,
'job_name': "job_name_1"
}
valid_args_2 = {
'time_start': 0.2,
'duration': 0.4,
'ncpus': 2,
'nnodes': 2,
'timesignals': ts_complete_set_2,
'job_name': "job_name_2"
}
# a model job that WILL NOT be picked by the algorithm..
job1 = ModelJob(**valid_args)
job1.label = "label_nottobepicked"
# a model job that WILL be picked by the algorithm..
job2 = ModelJob(**valid_args_2)
job2.label = "label_includeme"
# dummy workload with 20 jobs
np.random.seed(0)
jobs_all = []
for i in range(20):
# spawn a new job from either job1 or job2
if np.random.rand() < 0.5:
new_job = copy.deepcopy(job1)
else:
new_job = copy.deepcopy(job2)
# assign it a new label
jobs_all.append(new_job)
# create a workload out of all the jobs..
workload = Workload(jobs=jobs_all, tag="testing_workload")
# configure the splitter from user config
config_splitting = {
"type": "split",
"keywords_in": ["includeme"],
"keywords_out": ["excludeme"],
"split_by": "label",
"apply_to": ["testing_workload"],
"create_workload": "spawn_workload"
}
workloads = [workload]
splitter = WorkloadSplit(workloads)
splitter.apply(config_splitting)
wl_out = None
for wl in workloads:
if wl.tag == config_splitting["create_workload"]:
wl_out = wl
break
# make sure that we have created a workload as expected
self.assertTrue(wl_out is not None)
self.assertEqual(wl_out.tag, config_splitting["create_workload"])
# make sure that all the jobs have a label consistent with the filter
for j in wl_out.jobs:
self.assertTrue("includeme" in j.label
and "excludeme" not in j.label)
