def init_app():
global _db, _logs_dir, _k8s, _scheduler
logger.info('configuration dump\n%s', config.dump_yaml())
if config.httpdb.db_type == 'sqldb':
logger.info('using SQLDB')
_db = SQLDB(config.httpdb.dsn)
else:
logger.info('using FileRunDB')
_db = FileRunDB(config.httpdb.dirpath)
_db.connect()
_logs_dir = Path(config.httpdb.logs_path)
try:
_k8s = K8sHelper()
except Exception:
pass
# @yaronha - Initialize here
task = periodic.Task()
periodic.schedule(task, 60)
_scheduler = Scheduler()
for data in _db.list_schedules():
if 'schedule' not in data:
logger.warning('bad scheduler data - %s', data)
continue
_submit(data)
async def startup_event():
logger.info("configuration dump", dumped_config=config.dump_yaml())
await _initialize_singletons()
# periodic cleanup is not needed if we're not inside kubernetes cluster
if get_k8s_helper(silent=True).is_running_inside_kubernetes_cluster():
_start_periodic_cleanup()
_start_periodic_runs_monitoring()
async def startup_event():
logger.info("configuration dump\n%s", config.dump_yaml())
initialize_singletons()
task = periodic.Task()
periodic.schedule(task, 60)
_reschedule_tasks()
_start_periodic_cleanup()
def main():
global _file_db
from mlrun.config import config
logger.info('configuration dump\n%s', config.dump_yaml())
_file_db = FileRunDB(config.httpdb.dirpath, '.yaml')
_file_db.connect()
app.run(
host='0.0.0.0',
port=config.httpdb.port,
debug=config.httpdb.debug,
)
async def startup_event():
logger.info("configuration dump\n%s", config.dump_yaml())
initialize_singletons()
# don't fail the app on re-scheduling failure
try:
task = periodic.Task()
periodic.schedule(task, 60)
_reschedule_tasks()
except Exception as exc:
logger.warning(f'Failed rescheduling tasks, err: {exc}')
_start_periodic_cleanup()
async def startup_event():
logger.info("configuration dump", dumped_config=config.dump_yaml())
loop = asyncio.get_running_loop()
# Using python 3.8 default instead of 3.7 one - max(1, os.cpu_count()) * 5 cause it's causing to high memory
# consumption - https://bugs.python.org/issue35279
# TODO: remove when moving to python 3.8
max_workers = config.httpdb.max_workers or min(32, os.cpu_count() + 4)
loop.set_default_executor(
concurrent.futures.ThreadPoolExecutor(max_workers=max_workers))
await _initialize_singletons()
# periodic cleanup is not needed if we're not inside kubernetes cluster
if get_k8s_helper(silent=True).is_running_inside_kubernetes_cluster():
_start_periodic_cleanup()
_start_periodic_runs_monitoring()
async def startup_event():
logger.info(
"configuration dump",
dumped_config=config.dump_yaml(),
version=mlrun.utils.version.Version().get(),
)
loop = asyncio.get_running_loop()
# Using python 3.8 default instead of 3.7 one - max(1, os.cpu_count()) * 5 cause it's causing to high memory
# consumption - https://bugs.python.org/issue35279
# TODO: remove when moving to python 3.8
max_workers = config.httpdb.max_workers or min(32, os.cpu_count() + 4)
loop.set_default_executor(
concurrent.futures.ThreadPoolExecutor(max_workers=max_workers))
initialize_logs_dir()
initialize_db()
if config.httpdb.state == mlrun.api.schemas.APIStates.online:
await move_api_to_online()