def run_event_server(ftrack_url, ftrack_user, ftrack_api_key,
ftrack_events_path, no_stored_credentials,
store_credentials, legacy, clockify_api_key,
clockify_workspace):
if not no_stored_credentials:
cred = credentials.get_credentials(ftrack_url)
username = cred.get('username')
api_key = cred.get('api_key')
if clockify_workspace and clockify_api_key:
os.environ["CLOCKIFY_WORKSPACE"] = clockify_workspace
os.environ["CLOCKIFY_API_KEY"] = clockify_api_key
# Check url regex and accessibility
ftrack_url = check_ftrack_url(ftrack_url)
if not ftrack_url:
print('Exiting! < Please enter Ftrack server url >')
return 1
# Validate entered credentials
if not validate_credentials(ftrack_url, username, api_key):
print('Exiting! < Please enter valid credentials >')
return 1
if store_credentials:
credentials.save_credentials(username, api_key, ftrack_url)
# Set Ftrack environments
os.environ["FTRACK_SERVER"] = ftrack_url
os.environ["FTRACK_API_USER"] = username
os.environ["FTRACK_API_KEY"] = api_key
# TODO This won't work probably
if ftrack_events_path:
if isinstance(ftrack_events_path, (list, tuple)):
ftrack_events_path = os.pathsep.join(ftrack_events_path)
os.environ["FTRACK_EVENTS_PATH"] = ftrack_events_path
if legacy:
return legacy_server(ftrack_url)
return main_loop(ftrack_url)
def main(argv):
'''
There are 4 values neccessary for event server:
1.) Ftrack url - "studio.ftrackapp.com"
2.) Username - "my.username"
3.) API key - "apikey-long11223344-6665588-5565"
4.) Path/s to events - "X:/path/to/folder/with/events"
All these values can be entered with arguments or environment variables.
- arguments:
"-ftrackurl {url}"
"-ftrackuser {username}"
"-ftrackapikey {api key}"
"-ftrackeventpaths {path to events}"
- environment variables:
FTRACK_SERVER
FTRACK_API_USER
FTRACK_API_KEY
FTRACK_EVENTS_PATH
Credentials (Username & API key):
- Credentials can be stored for auto load on next start
- To *Store/Update* these values add argument "-storecred"
- They will be stored to appsdir file when login is successful
- To *Update/Override* values with enviromnet variables is also needed to:
- *don't enter argument for that value*
- add argument "-noloadcred" (currently stored credentials won't be loaded)
Order of getting values:
1.) Arguments are always used when entered.
- entered values through args have most priority! (in each case)
2.) Credentials are tried to load from appsdir file.
- skipped when credentials were entered through args or credentials
are not stored yet
- can be skipped with "-noloadcred" argument
3.) Environment variables are last source of values.
- will try to get not yet set values from environments
Best practice:
- set environment variables FTRACK_SERVER & FTRACK_EVENTS_PATH
- launch event_server_cli with args:
~/event_server_cli.py -ftrackuser "{username}" -ftrackapikey "{API key}" -storecred
- next time launch event_server_cli.py only with set environment variables
FTRACK_SERVER & FTRACK_EVENTS_PATH
'''
parser = argparse.ArgumentParser(description='Ftrack event server')
parser.add_argument("-ftrackurl",
type=str,
metavar='FTRACKURL',
help=("URL to ftrack server where events should handle"
" (default from environment: $FTRACK_SERVER)"))
parser.add_argument(
"-ftrackuser",
type=str,
help=("Username should be the username of the user in ftrack"
" to record operations against."
" (default from environment: $FTRACK_API_USER)"))
parser.add_argument("-ftrackapikey",
type=str,
help=("Should be the API key to use for authentication"
" (default from environment: $FTRACK_API_KEY)"))
parser.add_argument(
"-ftrackeventpaths",
nargs='+',
help=("List of paths where events are stored."
" (default from environment: $FTRACK_EVENTS_PATH)"))
parser.add_argument('-storecred',
help=("Entered credentials will be also stored"
" to apps dir for future usage"),
action="store_true")
parser.add_argument('-noloadcred',
help="Load creadentials from apps dir",
action="store_true")
parser.add_argument('-legacy',
help="Load creadentials from apps dir",
action="store_true")
ftrack_url = os.environ.get('FTRACK_SERVER')
username = os.environ.get('FTRACK_API_USER')
api_key = os.environ.get('FTRACK_API_KEY')
event_paths = os.environ.get('FTRACK_EVENTS_PATH')
kwargs, args = parser.parse_known_args(argv)
if kwargs.ftrackurl:
ftrack_url = kwargs.ftrackurl
if kwargs.ftrackeventpaths:
event_paths = kwargs.ftrackeventpaths
if not kwargs.noloadcred:
cred = credentials.get_credentials(ftrack_url)
username = cred.get('username')
api_key = cred.get('api_key')
if kwargs.ftrackuser:
username = kwargs.ftrackuser
if kwargs.ftrackapikey:
api_key = kwargs.ftrackapikey
legacy = kwargs.legacy
# Check url regex and accessibility
ftrack_url = check_ftrack_url(ftrack_url)
if not ftrack_url:
print('Exiting! < Please enter Ftrack server url >')
return 1
# Validate entered credentials
if not validate_credentials(ftrack_url, username, api_key):
print('Exiting! < Please enter valid credentials >')
return 1
# Process events path
event_paths, not_found = process_event_paths(event_paths)
if not_found:
print('WARNING: These paths were not found: {}'.format(str(not_found)))
if not event_paths:
if not_found:
print('ERROR: Any of entered paths is valid or can be accesible.')
else:
print('ERROR: Paths to events are not set. Exiting.')
return 1
if kwargs.storecred:
credentials.save_credentials(username, api_key, ftrack_url)
# Set Ftrack environments
os.environ["FTRACK_SERVER"] = ftrack_url
os.environ["FTRACK_API_USER"] = username
os.environ["FTRACK_API_KEY"] = api_key
os.environ["FTRACK_EVENTS_PATH"] = event_paths
if legacy:
return legacy_server(ftrack_url)
return main_loop(ftrack_url)
def main_loop(ftrack_url):
""" This is main loop of event handling.
Loop is handling threads which handles subprocesses of event storer and
processor. When one of threads is stopped it is tested to connect to
ftrack and mongo server. Threads are not started when ftrack or mongo
server is not accessible. When threads are started it is checked for socket
signals as heartbeat. Heartbeat must become at least once per 30sec
otherwise thread will be killed.
"""
os.environ["FTRACK_EVENT_SUB_ID"] = str(uuid.uuid1())
# Get mongo hostname and port for testing mongo connection
mongo_uri, mongo_port, database_name, collection_name = (
get_ftrack_event_mongo_info())
# Current file
file_path = os.path.dirname(os.path.realpath(__file__))
min_fail_seconds = 5
max_fail_count = 3
wait_time_after_max_fail = 10
# Threads data
storer_name = "StorerThread"
storer_port = 10001
storer_path = "{}/sub_event_storer.py".format(file_path)
storer_thread = None
storer_last_failed = datetime.datetime.now()
storer_failed_count = 0
processor_name = "ProcessorThread"
processor_port = 10011
processor_path = "{}/sub_event_processor.py".format(file_path)
processor_thread = None
processor_last_failed = datetime.datetime.now()
processor_failed_count = 0
statuser_name = "StorerThread"
statuser_port = 10021
statuser_path = "{}/sub_event_status.py".format(file_path)
statuser_thread = None
statuser_last_failed = datetime.datetime.now()
statuser_failed_count = 0
ftrack_accessible = False
mongo_accessible = False
printed_ftrack_error = False
printed_mongo_error = False
# stop threads on exit
# TODO check if works and args have thread objects!
def on_exit(processor_thread, storer_thread, statuser_thread):
if processor_thread is not None:
processor_thread.stop()
processor_thread.join()
processor_thread = None
if storer_thread is not None:
storer_thread.stop()
storer_thread.join()
storer_thread = None
if statuser_thread is not None:
statuser_thread.stop()
statuser_thread.join()
statuser_thread = None
atexit.register(on_exit,
processor_thread=processor_thread,
storer_thread=storer_thread,
statuser_thread=statuser_thread)
system_name, pc_name = platform.uname()[:2]
host_name = socket.gethostname()
main_info = {
"created_at": datetime.datetime.now().strftime("%Y.%m.%d %H:%M:%S"),
"Username": getpass.getuser(),
"Host Name": host_name,
"Host IP": socket.gethostbyname(host_name)
}
main_info_str = json.dumps(main_info)
# Main loop
while True:
# Check if accessible Ftrack and Mongo url
if not ftrack_accessible:
ftrack_accessible = check_ftrack_url(ftrack_url)
if not mongo_accessible:
mongo_accessible = check_mongo_url(mongo_uri, mongo_port)
# Run threads only if Ftrack is accessible
if not ftrack_accessible or not mongo_accessible:
if not mongo_accessible and not printed_mongo_error:
print("Can't access Mongo {}".format(mongo_uri))
if not ftrack_accessible and not printed_ftrack_error:
print("Can't access Ftrack {}".format(ftrack_url))
if storer_thread is not None:
storer_thread.stop()
storer_thread.join()
storer_thread = None
if processor_thread is not None:
processor_thread.stop()
processor_thread.join()
processor_thread = None
printed_ftrack_error = True
printed_mongo_error = True
time.sleep(1)
continue
printed_ftrack_error = False
printed_mongo_error = False
# ====== STATUSER =======
if statuser_thread is None:
if statuser_failed_count < max_fail_count:
statuser_thread = socket_thread.StatusSocketThread(
statuser_name, statuser_port, statuser_path,
[main_info_str])
statuser_thread.start()
elif statuser_failed_count == max_fail_count:
print(("Statuser failed {}times in row"
" I'll try to run again {}s later").format(
str(max_fail_count), str(wait_time_after_max_fail)))
statuser_failed_count += 1
elif ((datetime.datetime.now() - statuser_last_failed).seconds >
wait_time_after_max_fail):
statuser_failed_count = 0
# If thread failed test Ftrack and Mongo connection
elif not statuser_thread.isAlive():
statuser_thread.join()
statuser_thread = None
ftrack_accessible = False
mongo_accessible = False
_processor_last_failed = datetime.datetime.now()
delta_time = (_processor_last_failed -
statuser_last_failed).seconds
if delta_time < min_fail_seconds:
statuser_failed_count += 1
else:
statuser_failed_count = 0
statuser_last_failed = _processor_last_failed
elif statuser_thread.stop_subprocess:
print("Main process was stopped by action")
on_exit(processor_thread, storer_thread, statuser_thread)
os.kill(os.getpid(), signal.SIGTERM)
return 1
# ====== STORER =======
# Run backup thread which does not requeire mongo to work
if storer_thread is None:
if storer_failed_count < max_fail_count:
storer_thread = socket_thread.SocketThread(
storer_name, storer_port, storer_path)
storer_thread.start()
elif storer_failed_count == max_fail_count:
print(("Storer failed {}times I'll try to run again {}s later"
).format(str(max_fail_count),
str(wait_time_after_max_fail)))
storer_failed_count += 1
elif ((datetime.datetime.now() - storer_last_failed).seconds >
wait_time_after_max_fail):
storer_failed_count = 0
# If thread failed test Ftrack and Mongo connection
elif not storer_thread.isAlive():
if storer_thread.mongo_error:
raise MongoPermissionsError()
storer_thread.join()
storer_thread = None
ftrack_accessible = False
mongo_accessible = False
_storer_last_failed = datetime.datetime.now()
delta_time = (_storer_last_failed - storer_last_failed).seconds
if delta_time < min_fail_seconds:
storer_failed_count += 1
else:
storer_failed_count = 0
storer_last_failed = _storer_last_failed
# ====== PROCESSOR =======
if processor_thread is None:
if processor_failed_count < max_fail_count:
processor_thread = socket_thread.SocketThread(
processor_name, processor_port, processor_path)
processor_thread.start()
elif processor_failed_count == max_fail_count:
print(("Processor failed {}times in row"
" I'll try to run again {}s later").format(
str(max_fail_count), str(wait_time_after_max_fail)))
processor_failed_count += 1
elif ((datetime.datetime.now() - processor_last_failed).seconds >
wait_time_after_max_fail):
processor_failed_count = 0
# If thread failed test Ftrack and Mongo connection
elif not processor_thread.isAlive():
if processor_thread.mongo_error:
raise Exception(
"Exiting because have issue with acces to MongoDB")
processor_thread.join()
processor_thread = None
ftrack_accessible = False
mongo_accessible = False
_processor_last_failed = datetime.datetime.now()
delta_time = (_processor_last_failed -
processor_last_failed).seconds
if delta_time < min_fail_seconds:
processor_failed_count += 1
else:
processor_failed_count = 0
processor_last_failed = _processor_last_failed
if statuser_thread is not None:
statuser_thread.set_process("storer", storer_thread)
statuser_thread.set_process("processor", processor_thread)
time.sleep(1)
def legacy_server(ftrack_url):
# Current file
file_path = os.path.dirname(os.path.realpath(__file__))
min_fail_seconds = 5
max_fail_count = 3
wait_time_after_max_fail = 10
subproc = None
subproc_path = "{}/sub_legacy_server.py".format(file_path)
subproc_last_failed = datetime.datetime.now()
subproc_failed_count = 0
ftrack_accessible = False
printed_ftrack_error = False
while True:
if not ftrack_accessible:
ftrack_accessible = check_ftrack_url(ftrack_url)
# Run threads only if Ftrack is accessible
if not ftrack_accessible and not printed_ftrack_error:
print("Can't access Ftrack {} <{}>".format(
ftrack_url, str(datetime.datetime.now())))
if subproc is not None:
if subproc.poll() is None:
subproc.terminate()
subproc = None
printed_ftrack_error = True
time.sleep(1)
continue
printed_ftrack_error = False
if subproc is None:
if subproc_failed_count < max_fail_count:
subproc = subprocess.Popen(["python", subproc_path],
stdout=subprocess.PIPE)
elif subproc_failed_count == max_fail_count:
print(("Storer failed {}times I'll try to run again {}s later"
).format(str(max_fail_count),
str(wait_time_after_max_fail)))
subproc_failed_count += 1
elif ((datetime.datetime.now() - subproc_last_failed).seconds >
wait_time_after_max_fail):
subproc_failed_count = 0
# If thread failed test Ftrack and Mongo connection
elif subproc.poll() is not None:
subproc = None
ftrack_accessible = False
_subproc_last_failed = datetime.datetime.now()
delta_time = (_subproc_last_failed - subproc_last_failed).seconds
if delta_time < min_fail_seconds:
subproc_failed_count += 1
else:
subproc_failed_count = 0
subproc_last_failed = _subproc_last_failed
time.sleep(1)