The existing backend server implementations were inefficient since they
were sleeping for the full length of the timeouts rather than being woken when
there was data ready for them. It was assumed they would wake and perhaps did
when we forked processes directory but that is no longer the case.
This updates both the process and xmlrpc backends to wait using select(). This
does mean we need to pass the file descriptors to wait on from the internals
who know which these file descriptors are but this is a logical improvement.
Tests of a pathaolgical load on the process server of ~420 rapid tasks
executed on a server with BB_NUMBER_THREAD=48 went from a wall clock
measurement of the overall command execution time of 75s to a much more
reasonable 24s.
(Bitbake rev:
9bee497960889d9baa0a4284d79a384b18a8e826)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
if self.fakeworkerpipe:
self.fakeworkerpipe.read()
+ def active_fds(self):
+ fds = []
+ if self.workerpipe:
+ fds.append(self.workerpipe.input)
+ if self.fakeworkerpipe:
+ fds.append(self.fakeworkerpipe.input)
+ return fds
+
def check_stamp_task(self, task, taskname = None, recurse = False, cache = None):
def get_timestamp(f):
try:
(if the abort on failure configuration option isn't set)
"""
- retval = 0.5
+ retval = True
if self.state is runQueuePrepare:
self.rqexe = RunQueueExecuteDummy(self)
if self.stats.active > 0:
self.rq.read_workers()
- return 0.5
+ return self.rq.active_fds()
if len(self.failed_fnids) != 0:
self.rq.state = runQueueFailed
import signal
import sys
import time
+import select
from Queue import Empty
from multiprocessing import Event, Process, util, Queue, Pipe, queues
command = self.command_channel.recv()
self.runCommand(command)
- self.idle_commands(.1)
+ self.idle_commands(.1, [self.event_queue._reader, self.command_channel])
except Exception:
logger.exception('Running command %s', command)
self.cooker.stop()
self.idle_commands(.1)
- def idle_commands(self, delay):
+ def idle_commands(self, delay, fds = []):
nextsleep = delay
for function, data in self._idlefuns.items():
nextsleep = None
elif nextsleep is None:
continue
- elif retval < nextsleep:
- nextsleep = retval
+ else:
+ fds = fds + retval
except SystemExit:
raise
except Exception:
logger.exception('Running idle function')
if nextsleep is not None:
- time.sleep(nextsleep)
+ select.select(fds,[],[],nextsleep)
def runCommand(self, command):
"""
Serve Requests. Overloaded to honor a quit command
"""
self.quit = False
- self.timeout = 0 # Run Idle calls for our first callback
while not self.quit:
- #print "Idle queue length %s" % len(self._idlefuns)
- self.handle_request()
- #print "Idle timeout, running idle functions"
- nextsleep = None
+ fds = [self]
+ nextsleep = 0.1
for function, data in self._idlefuns.items():
try:
retval = function(self, data, False)
del self._idlefuns[function]
elif retval is True:
nextsleep = 0
- elif nextsleep is 0:
- continue
- elif nextsleep is None:
- nextsleep = retval
- elif retval < nextsleep:
- nextsleep = retval
+ else:
+ fds = fds + retval
except SystemExit:
raise
except:
import traceback
traceback.print_exc()
pass
- if nextsleep is None and len(self._idlefuns) > 0:
- nextsleep = 0
- self.timeout = nextsleep
+
+ socktimeout = self.socket.gettimeout() or nextsleep
+ socktimeout = min(socktimeout, nextsleep)
+ # Mirror what BaseServer handle_request would do
+ fd_sets = select.select(fds, [], [], socktimeout)
+ if fd_sets[0] and self in fd_sets[0]:
+ self._handle_request_noblock()
+
# Tell idle functions we're exiting
for function, data in self._idlefuns.items():
try: