Re-write downloader to be resilient to failures as much as possible

This makes the code crazy complicated and messy, but means we can be persistent about
not giving up, while still retrying at the same time, and trying multiple urls at once
until we find one that works.

See docstrings for a full discussion on some of the failures we're trying to work around.

downloader/downloader/__main__.py

@@ -1,3 +1,7 @@
+
+import gevent.monkey
+gevent.monkey.patch_all()
+
 import logging
 
 import argh

downloader/downloader/main.py

@@ -1,53 +1,456 @@
 
+import datetime
+import errno
 import hashlib
 import logging
 import os
-import time
+import random
+import uuid
 from base64 import b64encode
+from contextlib import contextmanager
+
+import dateutil.parser
+import gevent
+import gevent.event
+import requests
+from monotonic import monotonic
 
 import twitch
 
 
-def download_segments(base_dir, name, uri):
+class TimedOutError(Exception):
-	# NOTE: This function is not written to be resilient, do not rely on it.
+	pass
-	# There's no retry, no atomic rename for file writing, file name format is bad,
+
-	# skips segments with missing info, doesn't confirm lengths, other issues.
+
-	# Doesn't use monotonic time.
+@contextmanager
-	INTERVAL = 2
+def soft_hard_timeout(description, soft_timeout, hard_timeout, on_soft_timeout):
-	prev_segments = []
+	"""Context manager that wraps a piece of code in a pair of timeouts,
-	while True:
+	a "soft" timeout and a "hard" one. If the block does not complete before
-		last_check = time.time()
+	the soft timeout, the given on_soft_timeout() function is called in a new greenlet.
-		logging.info("Getting playlist")
+	If it doesn't complete before the hard timeout, a TimedOutError is raised.
-		playlist = twitch.get_media_playlist(uri)
+
-		for segment in playlist.segments:
+	Description is a short string, used for logging and error messages.
-			if segment.uri in prev_segments:
+
-				continue
+	A typical use-case is for the soft timeout to trigger some other code to begin
-			if not segment.date:
+	retrying, even as the original code continues to hold out in hope the call eventually
-				logging.warning("Segment has no date given, skipping")
+	succeeds.
+	"""
+	def dispatch_soft_timeout():
+		logging.warning("Hit soft timeout {}s while {}".format(soft_timeout, description))
+		on_soft_timeout()
+	soft_worker = gevent.spawn_later(soft_timeout, dispatch_soft_timeout)
+	error = TimedOutError("Timed out after {}s while {}".format(hard_timeout, description))
+	with gevent.Timeout(hard_timeout, error):
+		yield
+	# We didn't hard time out, if we also didn't soft timeout then cancel pending greenlet
+	if not soft_worker.started:
+		soft_worker.kill()
+
+
+def jitter(interval):
+	"""Apply some 'jitter' to an interval. This is a random +/- 10% change in order to
+	smooth out patterns and prevent everything from retrying at the same time.
+	"""
+	return interval * (0.9 + 0.2 * random.random())
+
+
+class StreamsManager(object):
+	"""Keeps track of what streams are being downloaded and the workers doing so.
+	Re-fetches master playlist when needed and starts new stream workers.
+	This is all to ensure that broken or bad media playlist urls are refreshed
+	in a timely manner.
+
+	The stream_workers map lists workers for each stream. Generally there should only be
+	one, but during switchover there may be 2 - one old one continuing to (try to) operate while
+	the second one confirms it's working. While trying to get a url working, it won't retry, it'll
+	just ask the manager immediately to create yet another new worker then quit.
+	When one successfully fetches a playlist for the first time, it marks all older
+	workers as able to shut down by calling manager.mark_working().
+
+	Creation of a new stream worker may be triggered by:
+	* An existing worker failing to refresh its playlist
+	* The master playlist finding a stream url it doesn't have already
+	* A worker is older than MAX_WORKER_AGE
+
+	The master playlist is only re-fetched as needed. Reasons to re-fetch:
+	* No url is known for at least one stream
+	* Someone has requested it (eg. because the previous attempt failed, or a worker needs a new url)
+	* A worker is older than MAX_WORKER_AGE
+	"""
+
+	FETCH_MIN_INTERVAL = 5
+	FETCH_SOFT_TIMEOUT = 5
+	FETCH_HARD_TIMEOUT = 30
+	MAX_WORKER_AGE = 20*60*60 # 20 hours, twitch's media playlist links expire after 24 hours
+
+	def __init__(self, channel, base_dir, qualities):
+		self.channel = channel
+		self.base_dir = base_dir
+		self.stream_workers = {name: [] for name in qualities + ["source"]} # {stream name: [workers]}
+		self.latest_urls = {} # {stream name: (fetch time, url)}
+		self.latest_urls_changed = gevent.event.Event() # set when latest_urls changes
+		self.refresh_needed = gevent.event.Event() # set to tell main loop to refresh now
+		self.stopping = gevent.event.Event() # set to tell main loop to stop
+
+	def mark_working(self, worker):
+		"""Notify the manager that the given worker is up and running,
+		and any older workers are safe to stop."""
+		workers = self.stream_workers[worker.stream]
+		if worker not in workers:
+			logging.warning("Worker {} called mark_working() but wasn't in known list: {}".format(worker, workers))
+			return
+		# stop everything older than given worker
+		for old in workers[:workers.index(worker)]:
+			old.stop()
+
+	def trigger_new_worker(self, worker):
+		"""Called when a worker decides a new worker for a stream is needed, eg. if it seems to be
+		failing. Causes a new worker with a fresh url to be created.
+		If worker's url is the same as the latest url, blocks until a new url has been fetched.
+		This only has effect if the worker is the current latest worker.
+		"""
+		workers = self.stream_workers[worker.stream]
+		if worker not in workers:
+			logging.warning("Worker {} called trigger_new_worker() but wasn't in known list: {}".format(worker, workers))
+			return
+		if worker is not workers[-1]:
+			logging.info("Ignoring request to start new worker for {} as old one is not latest".format(worker.stream))
+			return
+		logging.info("Starting new worker for {} as old one is failing", worker.stream)
+		self.wait_for_new_url(worker.stream, worker.url)
+		self.start_worker(worker.stream)
+		self.trigger_refresh()
+
+	def trigger_refresh(self):
+		"""Indicate that the master playlist needs to be re-fetched at the next available opportunity."""
+		self.refresh_needed.set()
+
+	def update_urls(self, time, new_urls):
+		"""Update the urls for streams, and notify anyone waiting"""
+		self.latest_urls.update({
+			name: (time, url)
+			for name, url in new_urls.items()
+		})
+		# set the old Event, and create a new one for any new waiters
+		self.latest_urls_changed.set()
+		self.latest_urls_changed = gevent.event.Event()
+
+	def wait_for_new_url(self, stream, old_url):
+		"""Trigger urls to be re-fetched, and block until a different one is received."""
+		while True:
+			if self.latest_urls[stream] != old_url:
+				return
+			self.trigger_refresh()
+			self.latest_urls_changed.wait()
+
+	def stop(self):
+		"""Shut down all workers and stop capturing stream."""
+		logging.info("Stopping streams manager")
+		self.stopping.set()
+
+	def start_worker(self, stream):
+		"""Start a new worker for given stream"""
+		url_time, url = self.latest_urls[stream]
+		worker = StreamWorker(self, stream, url, url_time)
+		self.stream_workers[stream].append(worker)
+		gevent.spawn(worker.run)
+
+	def fetch_latest(self):
+		"""Re-fetch master playlist and start new workers if needed"""
+		try:
+			# Fetch playlist. On soft timeout, retry.
+			logging.info("Fetching master playlist")
+			fetch_time = monotonic()
+			with soft_hard_timeout(self.FETCH_SOFT_TIMEOUT, self.FETCH_HARD_TIMEOUT, self.trigger_refresh):
+				master_playlist = twitch.get_master_playlist(self.channel)
+			new_urls = twitch.get_media_playlist_uris(master_playlist, self.qualities)
+			self.update_urls(fetch_time, new_urls)
+			for stream, workers in self.stream_workers.items():
+				# warn and retry if the url is missing
+				if stream not in new_urls:
+					logging.warning("Stream {} could not be found in latest master playlist, re-queueing refresh".format(stream))
+					self.trigger_refresh()
+				# is it newly found?
+				if not workers and stream in self.latest_urls:
+					logging.info("Starting new worker for {} as none exist".format(stream))
+					self.start_worker(stream)
+				latest_worker = workers[-1]
+				# is the old worker too old?
+				if latest_worker.age() > self.MAX_WORKER_AGE:
+					logging.info("Starting new worker for {} as the latest is too old ({}h)".format(stream, latest_worker.age() / 3600.))
+					self.start_worker(stream)
+		except Exception as e:
+			logging.exception("Failed to fetch master playlist")
+			# don't retry on hard timeout as we already retried on soft timeout
+			if not isinstance(e, TimedOutError):
+				self.trigger_refresh()
+
+	def run(self):
+		while not self.stopping.is_set():
+			times_to_max_age = [
+				self.MAX_WORKER_AGE - workers[-1].age()
+				for workers in self.stream_workers.values() if workers
+			]
+			time_to_next_max_age = min(times_to_max_age) if times_to_max_age else None
+			# wait until refresh triggered or next max age reached
+			self.refresh_needed.wait(time_to_next_max_age)
+			gevent.spawn(self.fetch_latest)
+			# wait min retry interval with jitter, unless we're stopping
+			self.stopping.wait(jitter(self.FETCH_MIN_INTERVAL))
+		logging.info("Stopping workers")
+		for workers in self.stream_workers.values():
+			for worker in workers:
+				worker.stop()
+			for worker in workers:
+				worker.done.wait()
+
+
+class StreamWorker(object):
+	"""Handles downloading segments for a particular media playlist.
+
+	If it fails to fetch its playlist (or hits the soft timeout), it tells the manager
+	to start a new worker with a new url.
+
+	If the url had been working (ie. that wasn't the first fetch), it will also stay alive and
+	attempt to use the old url until a new worker tells it to stand down,
+	or a 403 Forbidden is received (as this indicates the url is expired).
+
+	Since segment urls returned for a particular media playlist url are stable, we have an easier
+	time of managing downloading those:
+	* Every time a new URL is seen, a new SegmentGetter is created
+	* SegmentGetters will retry the same URL until they succeed, or get a 403 Forbidden indicating
+	  the url has expired.
+	"""
+
+	FETCH_SOFT_TIMEOUT = 5
+	FETCH_HARD_TIMEOUT = 90
+	FETCH_RETRY_INTERVAL = 0.5
+	FETCH_POLL_INTERVAL = 2
+
+	def __init__(self, manager, stream, url, url_time):
+		self.manager = manager
+		self.stream = stream
+		self.url = url
+		self.url_time = url_time
+		self.stopping = gevent.event.Event() # set to stop main loop
+		self.getters = {} # map from url to SegmentGetter
+		self.done = gevent.event.Event() # set when stopped and all getters are done
+
+	def age(self):
+		"""Return age of our url"""
+		return monotonic() - self.url_time
+
+	def stop(self):
+		"""Tell the worker to shut down"""
+		self.stopping.set()
+
+	def run(self):
+		try:
+			self._run()
+		except Exception:
+			logging.exception("Worker {} failed".format(self))
+			self.trigger_new_worker()
+		finally:
+			for getter in self.getters.values():
+				getter.done.wait()
+			self.done.set()
+			self.manager[self.stream].remove(self)
+
+	def trigger_new_worker(self):
+		self.manager.trigger_new_worker(self)
+
+	def wait(self, interval):
+		"""Wait for given interval with jitter, unless we're stopping"""
+		self.stopping.wait(jitter(interval))
+
+	def _run(self):
+		first = True
+		while not self.stopping.is_set():
+
+			logging.debug("{} getting media playlist {}".format(self, self.url))
+			try:
+				with soft_hard_timeout(self.FETCH_SOFT_TIMEOUT, self.FETCH_HARD_TIMEOUT, self.trigger_new_worker):
+					playlist = twitch.get_media_playlist(self.url)
+			except Exception as e:
+				logging.warning("{} failed to fetch media playlist {}".format(self, self.url), exc_info=True)
+				self.trigger_new_worker()
+				if first:
+					logging.debug("{} failed on first fetch, stopping".format(self))
+					self.stop()
+				elif isinstance(e, requests.HTTPError) and e.response and e.response.status_code == 403:
+					logging.debug("{} failed with 403 Forbidden, stopping".format(self))
+					self.stop()
+				self.wait(self.FETCH_RETRY_INTERVAL)
 				continue
-			segment_data = twitch.get_segment(segment.uri)
+
-			segment_hash = hashlib.sha256(segment_data).digest()
+			# We successfully got the playlist at least once
-			filename = os.path.join(
+			first = False
-				base_dir, name,
+
-				"{}-{}".format(segment.date, segment.duration),
+			# Start any new segment getters
-				"{}.ts".format(b64encode(segment_hash, "-_")),
+			date = None # tracks date in case some segment doesn't include it
-			)
+			for segment in playlist.segments:
-			logging.info("Writing segment {}".format(filename))
+				if segment.date:
-			if not os.path.exists(os.path.dirname(filename)):
+					date = dateutil.parser.parse(date)
-				os.makedirs(os.path.dirname(filename))
+				if segment.uri not in self.getters:
-			with open(filename, 'w') as f:
+					if date is None:
-				f.write(segment_data)
+						raise ValueError("Cannot determine date of segment")
-		prev_segments = [segment.uri for segment in playlist.segments]
+					self.getters[segment.uri] = SegmentGetter(self.manager.base_dir, self.stream, segment, date)
-		# Wait up to 2s until 2s after last_check
+					gevent.spawn(self.getters[segment.uri].run)
-		elapsed = time.time() - last_check
+				if date is not None:
-		time.sleep(max(0, INTERVAL - elapsed))
+					date += datetime.timedelta(seconds=segment.duration)
+
+			# Clean up any old segment getters
+			for url, getter in self.getters.items():
+				# If segment is done and wasn't in latest fetch
+				if getter.done.is_set() and not any(
+					segment.uri == url for segment in playlist.segments
+				):
+					del self.getters[url]
+
+			# Wait until next poll
+			self.wait(self.FETCH_POLL_INTERVAL)
+
+
+class SegmentGetter(object):
+	"""Fetches a segment and writes it to disk.
+	Retries until it succeeds, or gets a 403 Forbidden indicating
+	the url has expired.
+
+	Due to retries and multiple workers further up the stack, SegmentGetter needs to
+	assume there may be multiple getters trying to get the same segment. It handles this
+	by writing to a tempfile first, then atomically renaming to the computed filename
+	if it doesn't already exist.
+	"""
+	UNEXPECTED_FAILURE_RETRY = 0.5
+	FETCH_RETRY = 2
+	FETCH_HEADERS_SOFT_TIMEOUT = 5
+	FETCH_HEADERS_HARD_TIMEOUT = 30
+	FETCH_FULL_SOFT_TIMEOUT = 15
+	FETCH_FULL_HARD_TIMEOUT = 240
+
+	def __init__(self, base_dir, stream, segment, date):
+		self.segment = segment
+		self.date = date
+		self.prefix = self.make_path_prefix()
+		self.retry = None # Event, set to begin retrying
+		self.done = gevent.event.Event() # set when file exists or we give up
+
+	def run(self):
+		try:
+			while True:
+				try:
+					self._run()
+				except Exception:
+					logging.exception("Failure in SegmentGetter {}".format(self.segment))
+					gevent.sleep(jitter(self.UNEXPECTED_FAILURE_RETRY))
+				else:
+					break
+		finally:
+			self.done.set()
+
+	def _run(self):
+		while not self.exists():
+			self.retry = gevent.event.Event()
+			worker = gevent.spawn(self.get_segment)
+			# wait until worker succeeds/fails or retry is set
+			gevent.wait([worker, self.retry], count=1)
+			# If worker has returned, and return value is true, we're done
+			if worker.ready() and worker.value:
+				break
+			# if retry not set, wait for FETCH_RETRY first
+			self.retry.wait(jitter(self.FETCH_RETRY))
+
+	def make_path_prefix(self):
+		"""Generate leading part of filepath which doesn't change with the hash."""
+		return os.path.join(
+			self.base_dir,
+			self.stream,
+			self.date.strftime("%Y-%m-%dT%H"),
+			"{date}-{duration}".format(
+				date=self.date.strftime("%M:%S.%f"),
+				duration=self.segment.duration,
+			),
+		)
+
+	def make_path(self, type, hash=None):
+		"""Generate filepath for the segment.
+		Type may be:
+			full: Segment is complete. Hash is included.
+			partial: Segment is incomplete. Hash is included.
+			temp: Segment has not been downloaded yet. A random uuid is added.
+		"""
+		arg = str(uuid.uuid4()) if type == "temp" else b64encode(hash.digest(), "-_").rstrip("=")
+		return "{}-{}-{}.ts".format(self.prefix, type, arg)
+
+	def exists(self):
+		"""Look for an existing, full (non-partial) copy of this segment. Return bool."""
+		dirname = os.path.dirname(self.prefix)
+		try:
+			candidates = os.listdir(dirname)
+		except OSError as e:
+			# on ENOENT (doesn't exist), return []
+			if e.errno != errno.ENOENT:
+				raise
+			return []
+		full_prefix = "{}-full".format(self.prefix)
+		return any(candidate.startswith(full_prefix) for candidate in candidates)
+
+	def ensure_directory(self, path):
+		"""Create directory that contains path, as well as any parent directories,
+		if they don't already exist."""
+		dir_path = os.path.dirname(path)
+		if os.path.exists(dir_path):
+			return
+		self.ensure_directory(dir_path)
+		try:
+			os.mkdir(dir_path)
+		except OSError as e:
+			# Ignore if EEXISTS. This is needed to avoid a race if two getters run at once.
+			if e.errno != errno.EEXIST:
+				raise
+
+	def get_segment(self):
+		# save current value of self.retry so we can't set any later instance
+		# after a retry for this round has already occurred.
+		try:
+			self._get_segment()
+		except Exception:
+			logging.exception("Failed to get segment {}".format(self.segment))
+			return False
+		else:
+			return True
+
+	def _get_segment(self):
+		temp_path = self.make_path("temp")
+		hash = hashlib.sha256()
+		file_created = False
+		try:
+			with soft_hard_timeout(self.FETCH_FULL_SOFT_TIMEOUT, self.FETCH_FULL_HARD_TIMEOUT, self.retry.set):
+				with soft_hard_timeout(self.FETCH_HEADERS_SOFT_TIMEOUT, self.FETCH_HEADERS_HARD_TIMEOUT, self.retry.set):
+					resp = requests.get(self.segment.uri, stream=True)
+				if resp.status_code == 403:
+					logging.warning("Got 403 Forbidden for segment, giving up: {}".format(self.segment))
+					return
+				resp.raise_for_status()
+				self.ensure_directory(temp_path)
+				with open(temp_path, 'w') as f:
+					file_created = True
+					# We read chunk-wise in 8KiB chunks. Note that if the connection cuts halfway,
+					# we may lose part of the last chunk even though we did receive it.
+					# This is a small enough amount of data that we don't really care.
+					for chunk in resp.iter_content(8192):
+						f.write(chunk)
+						hash.update(chunk)
+		except Exception:
+			if file_created:
+				os.rename(temp_path, self.make_path("partial", hash))
+			raise
+		else:
+			os.rename(temp_path, self.make_path("full", hash))
 
 
 def main(channel, base_dir=".", qualities=""):
 	qualities = qualities.split(",") if qualities else []
-	master_playlist = twitch.get_master_playlist(channel)
+	manager = StreamsManager(channel, base_dir, qualities)
-	uris = twitch.get_media_playlist_uris(master_playlist, qualities)
+	logging.info("Starting up")
-	if len(uris) != 1:
+	manager.run()
-		raise NotImplementedError
-	(name, uri), = uris.items()
-	download_segments(base_dir, name, uri)

downloader/downloader/twitch.py

@@ -94,11 +94,3 @@ def get_media_playlist(uri):
 	resp = requests.get(uri)
 	resp.raise_for_status()
 	return hls_playlist.load(resp.text, base_uri=resp.url)
-
-
-def get_segment(uri):
-	# Future work: Make this stream directly to a file + hash it as we go,
-	# to avoid keeping it all in memory at once.
-	resp = requests.get(uri)
-	resp.raise_for_status()
-	return resp.content

downloader/setup.py

@@ -5,6 +5,10 @@ setup(
 	version = "0.0.0",
 	packages = find_packages(),
 	install_requires = [
+		"argh",
+		"dateutil",
+		"gevent",
+		"monotonic",
 		"requests",
 		"wubloader-common",
 	],