def test_two_asynch_running_retry_locks(self):
#--- get lock & hold it
cmd1 = '''%s/run_cletus_job_once.py \
--lock-wait 0 \
--post-lock-sleep 2 \
''' % test_path
print(cmd1)
self.c = envoy.connect(cmd1)
#---- ensure cmd1 locks file before cmd2 starts!
time.sleep(0.5)
#---- try to get lock, wait for it, get it
cmd2 = '''%s/run_cletus_job_once.py \
--lock-wait 3 \
--post-lock-sleep 0 \
''' % test_path
print(cmd2)
self.c2 = envoy.connect(cmd2)
#---- finish cmd2, then finish cmd1
self.c2.block()
self.c.block()
assert self.c2.status_code == 0 # locked
assert self.c.status_code == 0 # locked
def setUp(self):
global _httpbin
if not _httpbin:
self.httpbin = envoy.connect('gunicorn httpbin:app --bind=0.0.0.0:%s' % (PORT))
_httpbin = True
time.sleep(1)
def test_two_asynch_running(self):
#---- get lock & hold it
cmd1 = '''%s/run_cletus_job_once.py \
--lock-wait 0 \
--post-lock-sleep 3 \
''' % test_path
self.c = envoy.connect(cmd1)
cmd1_start_time = time.time()
#---- ensure cmd1 locks file before cmd2 starts!
time.sleep(0.5)
print('sleep for 0.5 seconds')
#---- try to get lock, fail, quit fast
cmd2 = '''%s/run_cletus_job_once.py \
--lock-wait 0 \
--post-lock-sleep 0 \
''' % test_path
self.c2 = envoy.connect(cmd2)
cmd2_start_time = time.time()
#---- finish cmd2, then finish cmd1
self.c2.block()
cmd2_dur = time.time() - cmd2_start_time
self.c.block()
cmd1_dur = time.time() - cmd1_start_time
assert self.c.status_code == 0 # locked
assert self.c2.status_code != 0 # not-locked
assert cmd2_dur < 2.0
print('\nfirst process info:')
print(cmd1)
print(self.c.std_out)
print(self.c.std_err)
print('cmd1 dur: %f' % cmd1_dur)
print('cmd1 lock status: %s' % self.c.status_code)
print('second process info:')
print(cmd2)
print(self.c2.std_out)
print(self.c2.std_err)
print('cmd2 dur: %f' % cmd2_dur)
print('cmd2 lock status: %s' % self.c2.status_code)
def setUp(self):
global _httpbin
if not _httpbin:
c = envoy.connect('gunicorn httpbin:app --bind=0.0.0.0:%s' % (PORT))
self.httpbin = c
_httpbin = True
def setUp(self):
global _httpbin
if not _httpbin:
c = envoy.connect('gunicorn httpbin:app --bind=0.0.0.0:%s' % (PORT))
self.httpbin = c
_httpbin = True
time.sleep(.1)
def setUp(self):
global _httpbin
if not _httpbin:
c = envoy.connect("gunicorn httpbin:app --bind=0.0.0.0:%s" % (PORT))
self.httpbin = c
_httpbin = True
time.sleep(1)
def test_single_asynch_process(self): # turned off for temp testing
""" Objective is to confirm that this method of
running the job will work correctly.
"""
cmd1 = '''%s/run_cletus_job_once.py \
--lock-wait 1 \
--post-lock-sleep 1 \
''' % test_path
#--------------------------------------------
# note that due to a bug in envoy, you must
# block before you can check the status_code
#-------------------------------------------
self.c = envoy.connect(cmd1)
self.c.block()
assert self.c.status_code == 0 # locked
def install(package, connect=False):
# Create temp git directory.
git_path = os.path.join(tempfile.mkdtemp(), 'repo')
os.makedirs(git_path)
os.chdir(git_path)
# Initialize the repo.
envoy.run('git init')
with open('requirements.txt', 'w') as f:
f.write(package)
envoy.run('git add requirements.txt')
envoy.run('git commit -m \'init\'')
# Create a new app.
print 'creating'
h_app = heroku.apps.add(stack='cedar')
print h_app.__dict__
print 'pushing'
cmd = 'git push https://{u}:{p}@code.heroku.com/{app}.git master'.format(
u=urllib.quote(HEROKU_EMAIL),
p=urllib.quote(HEROKU_PASS),
app=h_app.name
)
if not connect:
r = envoy.run(cmd)
h_app.destroy()
else:
r = envoy.connect(cmd)
return r
def test_status_code_none(self):
c = envoy.connect("sleep 5")
self.assertEqual(c.status_code, None)
def test_status_code(self):
c = envoy.connect("sleep 5")
self.assertEqual(c.status_code, None)
def test_input(self):
test_string = 'asdfQWER'
r = envoy.connect("cat | tr [:lower:] [:upper:]")
r.send(test_string)
self.assertEqual(r.std_out, test_string.upper())
self.assertEqual(r.status_code, 0)
def test_status_code_failure(self):
c = envoy.connect("sleeep 1")
c.block()
self.assertEqual(c.status_code, 127)
def test_status_code_success(self):
c = envoy.connect("sleep 1")
c.block()
self.assertEqual(c.status_code, 0)
#!/usr/bin/env python
from os import popen
from envoy import connect, run
bin="/Applications/Sublime Text 2.app/Contents/SharedSupport/bin/subl"
d="/Users/nordmenss/git/GISTS/4039460"
#connect(bin.replace(" ","\\ "))
#connect(bin.replace(" ","\\ ")+" -n "+d)
#run(bin.replace(" ","\\ ")," -n "+d)
#popen("/Applications/Sublime\\ Text\\ 2.app/Contents/SharedSupport/bin/subl -n /Users/nordmenss/git/GISTS/4039460")
print connect("say hi")
def test_input(self):
test_string = 'asdfQWER'
r = envoy.connect("cat | tr [:lower:] [:upper:]")
r.send(test_string)
self.assertEqual(r.std_out, test_string.upper())
self.assertEqual(r.status_code, 0)
def test_status_code_success(self):
c = envoy.connect("sleep 1")
time.sleep(2)
self.assertEqual(c.status_code, 0)
from time import sleep
import envoy
run = envoy.connect('make run')
print 'Running downloader...'
while True:
print 'Wating 200 seconds...'
sleep(200)
print 'Uploading...'
c = envoy.run('make upload')
print 'In sync!'
# print c.std_out
# print
def test_status_code_failure(self):
c = envoy.connect("sleep 1")
self.assertEqual(c.status_code, 127)
def test_status_code_success(self):
c = envoy.connect("sleep 1")
time.sleep(2)
self.assertEqual(c.status_code, 0)
def __aggregator_process_archiveitem(aggregator_archive_item, scheduler,
tmpdir, context):
import envoy
from django.template.loader import render_to_string
from webui.cnmain.utils import get_virtuoso
virtuoso_simple = get_virtuoso()
virtuoso_master = get_virtuoso('master')
loggy = local.logger
aggregator = aggregator_archive_item.aggregator
archive_item = aggregator_archive_item.archiveitem
#
# PART 1: generate XML file
#
loggy.debug("Processing " + unicode(archive_item))
output_filename = None
if not aggregator.silk_rule:
loggy.warning('No silk rule found, skipping')
scheduler.status = Scheduler.INCOMPLETE
else:
output_filename = os.path.join(tmpdir, archive_item.file_hash + '.nt')
conf_filename = os.path.join(tmpdir,
archive_item.file_hash + '_conf.xml')
silk_conf_xml = render_to_string(
'controller/aggregator/silk_rules.xml',
dict(context,
archive_item=archive_item,
output_filename=output_filename))
with open(conf_filename, 'w') as fconf:
fconf.write(silk_conf_xml)
#
# PART 2: execute SILK
#
loggy.info("Executing SILK on %s", unicode(archive_item))
result = envoy.connect(
'java -Xmx{} -DconfigFile={} -Dthreads={} '
'-cp "{}:{}/*" de.fuberlin.wiwiss.silk.Silk'.format(
settings.SILK_SINGLE_MACHINE_HEAP,
conf_filename,
settings.SILK_SINGLE_MACHINE_THREADS,
SILK_JAR_PATH,
SILK_LIB_PATH,
))
level = None
status = 0
titan_log_cnt = 0
# pylint: disable=W0212
while result._process.poll() is None:
line = result._process.stderr.readline()\
.strip().replace('%', '%%')
if not line:
continue
tmplevel = line.split(":", 1)[0]
if tmplevel in LEVEL_LIST:
level = tmplevel
if line.startswith("Exception in thread"):
level = "EXCEPTION"
if level == "EXCEPTION":
status = 2
loggy.error("S> " + line)
elif level in LEVEL_OUT:
status = 1
loggy.warn("S> " + line)
elif re.search(r"Finished writing \d+ entities", line) or \
re.search(r"Got \d+ vertices", line) or \
re.search(r"Wrote \d+ links", line):
loggy.info("S> " + line)
elif re.search(r"Getting data for vertices", line):
if titan_log_cnt % 200 == 0:
loggy.info("S> " + line)
titan_log_cnt += 1
# pylint: enable=W0212
if status:
loggy.error("SILK failed on %s", unicode(archive_item))
scheduler.status = Scheduler.FAIL
if status == 2:
return
else:
loggy.info("SILK executed successfully")
# loggy.debug("Generated file: %s", output_filename)
#
# PART 3: dump graph data
#
dump_dir = '{}/'.format(archive_item.file_hash)
loggy.info("Creating a dump of the namedgraph {}".format(
archive_item.datagraph_mapped_name))
error = virtuoso_simple.dump_graph(archive_item.datagraph_mapped_name,
dump_dir,
create_dir=True)
if error:
loggy.error("Dump failed:")
for line in error:
loggy.error(line)
raise Exception("Dump of the namedgraph failed: {}".format(error))
#
# PART 4: load graph data in the master virtuoso instance
#
# we are assuming that the two virtuoso are on the same machine
loggy.info("Loading dump in the master graph as {}".format(
archive_item.datagraph_mapped_name))
# clear the entire named database before ingesting the data
# since we're on titan we don't want this anymore
# virtuoso_master.clear(archive_item.datagraph_mapped_name)
# loggy.warning("Leaving data dump available for testing purposes")
# error = virtuoso_master.load_graphs(dump_dir, remove_dir=False)
error = virtuoso_master.load_graphs(dump_dir, remove_dir=True)
if error:
loggy.error("Load failed:")
if isinstance(error, basestring):
loggy.error(error)
else:
for line in error:
loggy.error(line)
raise Exception("Load of the namedgraph failed: {}".format(error))
if aggregator.silk_rule:
#
# PART 5: load SILK generated tuples
#
loggy.info("Loading SILK generated tuples")
virtuoso_master.ingest(
output_filename,
settings.TRIPLE_DATABASE['PREFIXES']['silk_graph'],
)
now = timezone.now()
aggregator_archive_item.last_workflow_success = now
if aggregator_archive_item.first_workflow_success is None:
aggregator_archive_item.first_workflow_success = now
aggregator_archive_item.save()
def __aggregator_process_archiveitem(
aggregator_archive_item, scheduler, tmpdir, context):
import envoy
from django.template.loader import render_to_string
from webui.cnmain.utils import get_virtuoso
virtuoso_simple = get_virtuoso()
virtuoso_master = get_virtuoso('master')
loggy = local.logger
aggregator = aggregator_archive_item.aggregator
archive_item = aggregator_archive_item.archiveitem
#
# PART 1: generate XML file
#
loggy.debug("Processing " + unicode(archive_item))
output_filename = None
if not aggregator.silk_rule:
loggy.warning('No silk rule found, skipping')
scheduler.status = Scheduler.INCOMPLETE
else:
output_filename = os.path.join(
tmpdir, archive_item.file_hash + '.nt'
)
conf_filename = os.path.join(
tmpdir, archive_item.file_hash + '_conf.xml'
)
silk_conf_xml = render_to_string(
'controller/aggregator/silk_rules.xml',
dict(context, archive_item=archive_item,
output_filename=output_filename)
)
with open(conf_filename, 'w') as fconf:
fconf.write(silk_conf_xml)
#
# PART 2: execute SILK
#
loggy.info("Executing SILK on %s", unicode(archive_item))
result = envoy.connect(
'java -Xmx{} -DconfigFile={} -Dthreads={} '
'-cp "{}:{}/*" de.fuberlin.wiwiss.silk.Silk'.format(
settings.SILK_SINGLE_MACHINE_HEAP,
conf_filename,
settings.SILK_SINGLE_MACHINE_THREADS,
SILK_JAR_PATH,
SILK_LIB_PATH,
)
)
level = None
status = 0
titan_log_cnt = 0
# pylint: disable=W0212
while result._process.poll() is None:
line = result._process.stderr.readline()\
.strip().replace('%', '%%')
if not line:
continue
tmplevel = line.split(":", 1)[0]
if tmplevel in LEVEL_LIST:
level = tmplevel
if line.startswith("Exception in thread"):
level = "EXCEPTION"
if level == "EXCEPTION":
status = 2
loggy.error("S> " + line)
elif level in LEVEL_OUT:
status = 1
loggy.warn("S> " + line)
elif re.search(r"Finished writing \d+ entities", line) or \
re.search(r"Got \d+ vertices", line) or \
re.search(r"Wrote \d+ links", line):
loggy.info("S> " + line)
elif re.search(r"Getting data for vertices", line):
if titan_log_cnt % 200 == 0:
loggy.info("S> " + line)
titan_log_cnt += 1
# pylint: enable=W0212
if status:
loggy.error("SILK failed on %s", unicode(archive_item))
scheduler.status = Scheduler.FAIL
if status == 2:
return
else:
loggy.info("SILK executed successfully")
# loggy.debug("Generated file: %s", output_filename)
#
# PART 3: dump graph data
#
dump_dir = '{}/'.format(archive_item.file_hash)
loggy.info("Creating a dump of the namedgraph {}".format(
archive_item.datagraph_mapped_name))
error = virtuoso_simple.dump_graph(
archive_item.datagraph_mapped_name, dump_dir, create_dir=True)
if error:
loggy.error("Dump failed:")
for line in error:
loggy.error(line)
raise Exception("Dump of the namedgraph failed: {}".format(
error
))
#
# PART 4: load graph data in the master virtuoso instance
#
# we are assuming that the two virtuoso are on the same machine
loggy.info("Loading dump in the master graph as {}".format(
archive_item.datagraph_mapped_name))
# clear the entire named database before ingesting the data
# since we're on titan we don't want this anymore
# virtuoso_master.clear(archive_item.datagraph_mapped_name)
# loggy.warning("Leaving data dump available for testing purposes")
# error = virtuoso_master.load_graphs(dump_dir, remove_dir=False)
error = virtuoso_master.load_graphs(dump_dir, remove_dir=True)
if error:
loggy.error("Load failed:")
if isinstance(error, basestring):
loggy.error(error)
else:
for line in error:
loggy.error(line)
raise Exception("Load of the namedgraph failed: {}".format(
error
))
if aggregator.silk_rule:
#
# PART 5: load SILK generated tuples
#
loggy.info("Loading SILK generated tuples")
virtuoso_master.ingest(
output_filename,
settings.TRIPLE_DATABASE['PREFIXES']['silk_graph'],
)
now = timezone.now()
aggregator_archive_item.last_workflow_success = now
if aggregator_archive_item.first_workflow_success is None:
aggregator_archive_item.first_workflow_success = now
aggregator_archive_item.save()
import envoy
import time
r = envoy.connect('python /Users/timbueno/Projects/Intellibrella/RaspberryPi/Other/subprocesses/process2.py cool')
run = True
count = 0
while run:
print 'Doing other things\n'
if count == 10:
r.kill()
print 'Killed the process...\n'
if count == 20:
run = False
count = count+1
time.sleep(1)
print 'The program has run its course'
def launch(self):
#if not os.path.exists(self.renderfilename):
self.render()
self.evince = envoy.connect('evince {}'.format(self.renderfilename))
