Update the worker code to better handle exceptions, fix the utcdate issue and make sure we send the proper retry. Also updates notification workers to send JobExceptions rather than returning true or false

This commit is contained in:
Joseph Schorr 2014-09-22 12:52:57 -04:00
parent 8c00eabedd
commit f23038c6ee
6 changed files with 62 additions and 60 deletions

View file

@ -63,11 +63,12 @@ class WorkerStatusHandler(BaseHTTPRequestHandler):
class Worker(object):
def __init__(self, queue, poll_period_seconds=30, reservation_seconds=300,
watchdog_period_seconds=60):
watchdog_period_seconds=60, retry_after_seconds=300):
self._sched = BackgroundScheduler()
self._poll_period_seconds = poll_period_seconds
self._reservation_seconds = reservation_seconds
self._watchdog_period_seconds = watchdog_period_seconds
self._retry_after_seconds = retry_after_seconds
self._stop = Event()
self._terminated = Event()
self._queue = queue
@ -103,7 +104,8 @@ class Worker(object):
try:
self.watchdog()
except WorkerUnhealthyException as exc:
logger.error('The worker has encountered an error via watchdog and will not take new jobs: %s' % exc.message)
logger.error('The worker has encountered an error via watchdog and will not take new jobs')
logger.error(exc.message)
self.mark_current_incomplete(restore_retry=True)
self._stop.set()
@ -111,7 +113,7 @@ class Worker(object):
logger.debug('Getting work item from queue.')
with self._current_item_lock:
self.current_queue_item = self._queue.get()
self.current_queue_item = self._queue.get(processing_time=self._reservation_seconds)
while True:
# Retrieve the current item in the queue over which to operate. We do so under
@ -129,12 +131,14 @@ class Worker(object):
self.process_queue_item(job_details)
self.mark_current_complete()
except JobException:
except JobException as jex:
logger.warning('An error occurred processing request: %s', current_queue_item.body)
logger.warning('Job exception: %s' % jex)
self.mark_current_incomplete(restore_retry=False)
except WorkerUnhealthyException as exc:
logger.error('The worker has encountered an error via the job and will not take new jobs: %s' % exc.message)
logger.error('The worker has encountered an error via the job and will not take new jobs')
logger.error(exc.message)
self.mark_current_incomplete(restore_retry=True)
self._stop.set()
@ -190,7 +194,8 @@ class Worker(object):
def mark_current_incomplete(self, restore_retry=False):
with self._current_item_lock:
if self.current_queue_item is not None:
self._queue.incomplete(self.current_queue_item, restore_retry=restore_retry)
self._queue.incomplete(self.current_queue_item, restore_retry=restore_retry,
retry_after=self._retry_after_seconds)
self.current_queue_item = None
def mark_current_complete(self):