def pause_pipeline(self, run_id, user):
"""
Interrupt pipeline by sending signal to corresponding worker's children
"""
pid = self.pids.get(run_id)
if pid:
pretty_print("Pausing pipeline: id=%d, user=%s" % (run_id, user))
try:
parent = psutil.Process(pid)
children = parent.children(recursive=True)
for child in children:
run_as(cmd=['kill', child.pid], user=user)
except psutil.NoSuchProcess:
pretty_print("Error pausing pipeline: no process with ID %d" % int(pid))
else:
pretty_print("Error pausing pipeline: ID %d not found" % run_id)
def resume(user, cfg, run_id, pids):
"""
Resumes as user 'user' a pipeline defined by the given config
Returns exit code, stdout, and stderr.
"""
pids[run_id] = mp.current_process().pid
cmd = [which('np_submit.py'), cfg]
(ec, err, out) = run_as(cmd=cmd, user=user)
if ec == 0:
return (err, out)
else:
raise Exception('Unable to execute cmd %s:\n %s' % (cmd, err))
def delete(user, output_dir, work_dir):
"""
Deletes the output directory of a pipeline.
Returns exit code, stdout, and stderr.
"""
if not work_dir:
cmd = ['rm', '-rf', output_dir]
else:
cmd = ['rm', '-rf', work_dir]
if work_dir != output_dir:
cmd.extend([';', 'rm', '-rf', output_dir])
(ec, err, out) = run_as(cmd, user=user)
out = ' '.join(cmd)+'\n'+out
if ec == 0:
return (err, out)
else:
raise Exception('Unable to execute cmd %s:\n %s' % (cmd, err))
def submit(config, user, run_id, pids):
"""
Submits pipeline defined by 'config' as user 'user'.
Dumps the config in a temp. file that is removed after succesful completion.
Returns exit code, stdout, and stderr.
"""
pids[run_id] = mp.current_process().pid
(fd, tmp_cfg) = tempfile.mkstemp(prefix='pypers_', suffix='.cfg', text=True)
os.fchmod(fd, 0644)
with os.fdopen(fd, 'w') as fh:
json.dump(config, fh)
cmd = [which('np_submit.py'), '-i', tmp_cfg]
(ec, err, out) = run_as(cmd=cmd, user=user)
if ec == 0:
os.unlink(tmp_cfg)
return (err, out)
else:
raise Exception('Unable to execute cmd %s:\n%s\n%s' % (cmd, err, out))
args = parser.parse_args()
user = getpass.getuser()
# create a step
step = Step.load_step(args.cfg_file)
if not os.path.exists(step.output_dir):
os.makedirs(step.output_dir, 0775)
# remove existing files, except step config and condor files
full_list = glob.glob(step.output_dir + "/*")
regex = re.compile("(job\.*|condor\.*|.*\.cfg)")
to_remove = filter(lambda f: not regex.search(f), full_list)
for entry in to_remove:
cmd = ['rm', '-rvf', entry]
(ec, err, out) = run_as(cmd=cmd, user=user)
if ec:
print "WARNING: failed to remove file %s: %s, %s" % (entry, err, out)
else:
print "Removed %s" % entry
# launch watchdog
mypid = os.getpid()
cmd = [WATCHDOG_EXE, str(mypid)]
watchlog = os.path.join(step.output_dir, 'watchdog.log')
wp = subprocess.Popen(cmd, stdout=open(watchlog,'w'), stderr=open(watchlog,'a'))
# run step
step.run()
# stop watchdog
def exec_monitoring(self):
"""
Check if all the flow cell IDs have been demultiplexed
For each flow cell which has not been demultiplexed,
then the demultiplexing pipeline is submitted to the cluster
"""
#Create a dictionary with {"Fw cell ID" : "path"}
fw_cell_dirs = {}
missing_ss_list = []
for hiseq_dir in self.hiseq_dirs:
#Parse all the hiseq dirs and create a list of data directories
#Only the directorise with the "RTAComplete.txt" file are considered
for fwcell in os.listdir(hiseq_dir):
fwcell_path = os.path.join(hiseq_dir, fwcell)
if (re.search(".+_.+_.+_.+", fwcell) \
and "Temp" not in fwcell \
and os.path.exists(os.path.join(fwcell_path, "RTAComplete.txt"))):
ss_found = False
#search for the sample sheet in the fwcell_path
for filename in os.listdir(fwcell_path):
if ("SampleSheet" in filename) and (".csv" in filename):
ss_found = True
break
if ss_found:
fw_cell_dirs[fwcell] = os.path.join(hiseq_dir, fwcell)
#otherwise add the directory to the list of missing sample sheet
else:
missing_ss_list.append(os.path.join(hiseq_dir, fwcell))
#log all the missing sample sheets detected
if missing_ss_list:
print ("******************************************************")
for missing_ss in missing_ss_list:
print ("Missing sample sheet in %s "% missing_ss)
#create a set for the hiseq dirs and a set for the demultiplexed dirs
hiseq_flow_cells = set([key for key in fw_cell_dirs])
demu_flow_cells = set(os.listdir(self.demu_dir))
if not hiseq_flow_cells.issubset(demu_flow_cells):
#get the difference
fwcell_diff = hiseq_flow_cells.difference(demu_flow_cells)
if fwcell_diff:
for fwcell_id in fwcell_diff:
submit_cmd = which('np_submit.py')
cmd = [
submit_cmd,
pipeline_names['demultiplexing'],
'pipeline.output_dir=%s' % os.path.join(self.demu_dir, fwcell_id),
'pipeline.project_name=Demux',
'pipeline.description=Demultiplexing',
'steps.inputs.input_dir=%s' % fw_cell_dirs[fwcell_id]
]
run_as(cmd=cmd, user=self.user)
print("******************************************************")
print(" %s Queued demux with:" % time.ctime())
print(" Input dir : %s" % fw_cell_dirs[fwcell_id])
print(" Output dir : %s" % os.path.join(self.demu_dir, fwcell_id))
print(" Cmd : %s" % ' '.join(cmd))
print("******************************************************")
args = parser.parse_args()
user = getpass.getuser()
# create a step
step = Step.load_step(args.cfg_file)
if not os.path.exists(step.output_dir):
os.makedirs(step.output_dir, 0775)
# remove existing files, except step config and condor files
full_list = glob.glob(step.output_dir + "/*")
regex = re.compile("(job\.*|condor\.*|.*\.cfg)")
to_remove = filter(lambda f: not regex.search(f), full_list)
for entry in to_remove:
cmd = ['rm', '-rvf', entry]
(ec, err, out) = run_as(cmd=cmd, user=user)
if ec:
print "WARNING: failed to remove file %s: %s, %s" % (entry, err,
out)
else:
print "Removed %s" % entry
# launch watchdog
mypid = os.getpid()
cmd = [WATCHDOG_EXE, str(mypid)]
watchlog = os.path.join(step.output_dir, 'watchdog.log')
wp = subprocess.Popen(cmd,
stdout=open(watchlog, 'w'),
stderr=open(watchlog, 'a'))
# run step
def exec_monitoring(self):
"""
Check if all the flow cell IDs have been demultiplexed
For each flow cell which has not been demultiplexed,
then the demultiplexing pipeline is submitted to the cluster
"""
#Create a dictionary with {"Fw cell ID" : "path"}
fw_cell_dirs = {}
missing_ss_list = []
for hiseq_dir in self.hiseq_dirs:
#Parse all the hiseq dirs and create a list of data directories
#Only the directorise with the "RTAComplete.txt" file are considered
for fwcell in os.listdir(hiseq_dir):
fwcell_path = os.path.join(hiseq_dir, fwcell)
if (re.search(".+_.+_.+_.+", fwcell) \
and "Temp" not in fwcell \
and os.path.exists(os.path.join(fwcell_path, "RTAComplete.txt"))):
ss_found = False
#search for the sample sheet in the fwcell_path
for filename in os.listdir(fwcell_path):
if ("SampleSheet" in filename) and (".csv"
in filename):
ss_found = True
break
if ss_found:
fw_cell_dirs[fwcell] = os.path.join(hiseq_dir, fwcell)
#otherwise add the directory to the list of missing sample sheet
else:
missing_ss_list.append(os.path.join(hiseq_dir, fwcell))
#log all the missing sample sheets detected
if missing_ss_list:
print("******************************************************")
for missing_ss in missing_ss_list:
print("Missing sample sheet in %s " % missing_ss)
#create a set for the hiseq dirs and a set for the demultiplexed dirs
hiseq_flow_cells = set([key for key in fw_cell_dirs])
demu_flow_cells = set(os.listdir(self.demu_dir))
if not hiseq_flow_cells.issubset(demu_flow_cells):
#get the difference
fwcell_diff = hiseq_flow_cells.difference(demu_flow_cells)
if fwcell_diff:
for fwcell_id in fwcell_diff:
submit_cmd = which('np_submit.py')
cmd = [
submit_cmd, pipeline_names['demultiplexing'],
'pipeline.output_dir=%s' %
os.path.join(self.demu_dir, fwcell_id),
'pipeline.project_name=Demux',
'pipeline.description=Demultiplexing',
'steps.inputs.input_dir=%s' % fw_cell_dirs[fwcell_id]
]
run_as(cmd=cmd, user=self.user)
print(
"******************************************************"
)
print(" %s Queued demux with:" % time.ctime())
print(" Input dir : %s" % fw_cell_dirs[fwcell_id])
print(" Output dir : %s" %
os.path.join(self.demu_dir, fwcell_id))
print(" Cmd : %s" % ' '.join(cmd))
print(
"******************************************************"
)