backfiller: Allow multiple concurrent segment downloads

This will signifigantly increase throughput when downloading
large ranges of segments.

The max concurrency is exposed as a cli arg.

We also slightly modify the logged info, so it reports segments downloaded,
not just number of missing segments (which we might skip downloading for various reasons).

backfiller/backfiller/main.py

@@ -12,6 +12,7 @@ import uuid
 
 import argh
 import gevent.backdoor
+import gevent.pool
 import prometheus_client as prom
 import requests
 
@@ -142,7 +143,8 @@ class BackfillerManager(object):
 	NODE_INTERVAL = 300 #seconds between updating list of nodes
 
 	def __init__(self, base_dir, stream, variants, static_nodes=[], start=None,
-			run_once=False, node_file=None, node_database=None, localhost=None):
+			run_once=False, node_file=None, node_database=None, localhost=None,
+			download_concurrency=5):
 		"""Constructor for BackfillerManager.
 		Creates a manager for a given stream with specified variants. If
 		static_nodes is None, manager"""
@@ -157,6 +159,7 @@ class BackfillerManager(object):
 		if self.node_database is not None:
 			self.db_manager =  database.DBManager(dsn=self.node_database)
 		self.localhost = localhost
+		self.download_concurrency = download_concurrency
 		self.stopping = gevent.event.Event()
 		self.logger = logging.getLogger("BackfillerManager({})".format(stream))
 		self.workers = {} # {node url: worker}
@@ -282,6 +285,7 @@ class BackfillerWorker(object):
 		self.logger = manager.logger.getChild('BackfillerWorker({})'.format(node))
 		self.base_dir = manager.base_dir
 		self.node = node
+		self.download_concurrency = manager.download_concurrency
 		self.stream = manager.stream
 		self.variants = manager.variants
 		self.start = manager.start
@@ -312,7 +316,7 @@ class BackfillerWorker(object):
 	
 			for hour in hours:
 	
-				self.logger.info('Backfilling {}/{}'.format(variant, hour))
+				self.logger.debug('Backfilling {}/{}'.format(variant, hour))
 	
 				local_segments = set(list_local_segments(self.base_dir, self.stream, variant, hour))
 				remote_segments = set(list_remote_segments(self.node, self.stream, variant, hour))
@@ -321,6 +325,9 @@ class BackfillerWorker(object):
 				# randomise the order of the segments to reduce the chance that
 				# multiple workers request the same segment at the same time
 				random.shuffle(missing_segments)
+
+				pool = gevent.pool.Pool(self.download_concurrency)
+				workers = []
 	
 				for missing_segment in missing_segments:
 	
@@ -347,8 +354,18 @@ class BackfillerWorker(object):
 						self.logger.debug('Skipping {} as too recent'.format(path))
 						continue
 	
-					get_remote_segment(self.base_dir, self.node, self.stream, variant, hour, missing_segment)
-				self.logger.info('{} segments in {}/{} backfilled'.format(len(missing_segments), variant, hour))
+					# start segment as soon as a pool slot opens up, then track it in workers
+					workers.append(pool.spawn(
+						get_remote_segment,
+						self.base_dir, self.node, self.stream, variant, hour, missing_segment
+					))
+
+				# verify that all the workers succeeded. if any failed, raise the exception from
+				# one of them arbitrarily.
+				for worker in workers:
+					worker.get() # re-raise error, if any
+
+				self.logger.info('{} segments in {}/{} backfilled'.format(len(workers), variant, hour))
 
 	def run(self):
 		self.logger.info('Starting')
@@ -388,9 +405,11 @@ class BackfillerWorker(object):
 @argh.arg('--node-file', help="Name of file listing nodes to backfill from. One node per line in the form NAME URI with whitespace only lines or lines starting with '#' ignored. If None (default) do not get nodes from a file.")
 @argh.arg('--node-database', help='Postgres conection string for database to fetch a list of nodes from. Either a space-separated list of key=value pairs, or a URI like: postgresql://USER:PASSWORD@HOST/DBNAME?KEY=VALUE . If None (default) do not get nodes from database.')
 @argh.arg('--localhost', help='Name of local machine. Used to prevent backfilling from itself. By default the result of socket.gethostname()')
+@argh.arg('--download-concurrency', help='Max number of concurrent segment downloads from a single node. Increasing this number may increase throughput but too high a value can overload the server or cause timeouts.')
 def main(streams, base_dir='.', variants='source', metrics_port=8002,
 	static_nodes='', backdoor_port=0, start=None, run_once=False,
-	node_file=None, node_database=None, localhost=socket.gethostname()):
+	node_file=None, node_database=None, localhost=socket.gethostname(),
+	download_concurrency=5):
 	"""Backfiller service."""
 
 	variants = variants.split(',') if variants else []
@@ -414,7 +433,7 @@ def main(streams, base_dir='.', variants='source', metrics_port=8002,
 	workers = []
 	for stream in streams:
 		logging.info('Starting backfilling {} with {} as variants to {}'.format(stream, ', '.join(variants), base_dir))
-		manager = BackfillerManager(base_dir, stream, variants, static_nodes, start, run_once, node_file, node_database, localhost)
+		manager = BackfillerManager(base_dir, stream, variants, static_nodes, start, run_once, node_file, node_database, localhost, download_concurrency)
 		managers.append(manager)
 		workers.append(gevent.spawn(manager.run))