/
arbiter.py
516 lines (444 loc) · 15.2 KB
/
arbiter.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
# -*- coding: utf-8 -
#
# This file is part of gunicorn released under the MIT license.
# See the NOTICE for more information.
from __future__ import with_statement
import errno
import os
import select
import signal
import sys
import time
import traceback
from gunicorn.errors import HaltServer
from gunicorn.pidfile import Pidfile
from gunicorn.sock import create_socket
from gunicorn import util
from gunicorn import __version__, SERVER_SOFTWARE
class Arbiter(object):
"""
Arbiter maintain the workers processes alive. It launches or
kills them if needed. It also manages application reloading
via SIGHUP/USR2.
"""
# A flag indicating if a worker failed to
# to boot. If a worker process exist with
# this error code, the arbiter will terminate.
WORKER_BOOT_ERROR = 3
START_CTX = {}
LISTENER = None
WORKERS = {}
PIPE = []
# I love dynamic languages
SIG_QUEUE = []
SIGNALS = map(
lambda x: getattr(signal, "SIG%s" % x),
"HUP QUIT INT TERM TTIN TTOU USR1 USR2 WINCH".split()
)
SIG_NAMES = dict(
(getattr(signal, name), name[3:].lower()) for name in dir(signal)
if name[:3] == "SIG" and name[3] != "_"
)
def __init__(self, app):
os.environ["SERVER_SOFTWARE"] = SERVER_SOFTWARE
self.setup(app)
self.pidfile = None
self.worker_age = 0
self.reexec_pid = 0
self.master_name = "Master"
# get current path, try to use PWD env first
try:
a = os.stat(os.environ['PWD'])
b = os.stat(os.getcwd())
if a.ino == b.ino and a.dev == b.dev:
cwd = os.environ['PWD']
else:
cwd = os.getcwd()
except:
cwd = os.getcwd()
args = sys.argv[:]
args.insert(0, sys.executable)
# init start context
self.START_CTX = {
"args": args,
"cwd": cwd,
0: sys.executable
}
def setup(self, app):
self.app = app
self.cfg = app.cfg
self.log = self.cfg.logger_class(app.cfg)
# reopen files
if 'GUNICORN_FD' in os.environ:
self.log.reopen_files()
self.address = self.cfg.address
self.num_workers = self.cfg.workers
self.debug = self.cfg.debug
self.timeout = self.cfg.timeout
self.proc_name = self.cfg.proc_name
self.worker_class = self.cfg.worker_class
if self.cfg.debug:
self.log.debug("Current configuration:")
for config, value in sorted(self.cfg.settings.iteritems()):
self.log.debug(" %s: %s", config, value.value)
if self.cfg.preload_app:
if not self.cfg.debug:
self.app.wsgi()
else:
self.log.warning("debug mode: app isn't preloaded.")
def start(self):
"""\
Initialize the arbiter. Start listening and set pidfile if needed.
"""
self.log.info("Starting gunicorn %s", __version__)
self.pid = os.getpid()
if self.cfg.pidfile is not None:
self.pidfile = Pidfile(self.cfg.pidfile)
self.pidfile.create(self.pid)
self.cfg.on_starting(self)
self.init_signals()
if not self.LISTENER:
self.LISTENER = create_socket(self.cfg, self.log)
self.log.debug("Arbiter booted")
self.log.info("Listening at: %s (%s)", self.LISTENER,
self.pid)
self.log.info("Using worker: %s",
self.cfg.settings['worker_class'].get())
self.cfg.when_ready(self)
def init_signals(self):
"""\
Initialize master signal handling. Most of the signals
are queued. Child signals only wake up the master.
"""
if self.PIPE:
map(os.close, self.PIPE)
self.PIPE = pair = os.pipe()
map(util.set_non_blocking, pair)
map(util.close_on_exec, pair)
self.log.close_on_exec()
map(lambda s: signal.signal(s, self.signal), self.SIGNALS)
signal.signal(signal.SIGCHLD, self.handle_chld)
def signal(self, sig, frame):
if len(self.SIG_QUEUE) < 5:
self.SIG_QUEUE.append(sig)
self.wakeup()
def run(self):
"Main master loop."
self.start()
util._setproctitle("master [%s]" % self.proc_name)
self.manage_workers()
while True:
try:
self.reap_workers()
sig = self.SIG_QUEUE.pop(0) if len(self.SIG_QUEUE) else None
if sig is None:
self.sleep()
self.murder_workers()
self.manage_workers()
continue
if sig not in self.SIG_NAMES:
self.log.info("Ignoring unknown signal: %s", sig)
continue
signame = self.SIG_NAMES.get(sig)
handler = getattr(self, "handle_%s" % signame, None)
if not handler:
self.log.error("Unhandled signal: %s", signame)
continue
self.log.info("Handling signal: %s", signame)
handler()
self.wakeup()
except StopIteration:
self.halt()
except KeyboardInterrupt:
self.halt()
except HaltServer, inst:
self.halt(reason=inst.reason, exit_status=inst.exit_status)
except SystemExit:
raise
except Exception:
self.log.info("Unhandled exception in main loop:\n%s",
traceback.format_exc())
self.stop(False)
if self.pidfile is not None:
self.pidfile.unlink()
sys.exit(-1)
def handle_chld(self, sig, frame):
"SIGCHLD handling"
self.wakeup()
def handle_hup(self):
"""\
HUP handling.
- Reload configuration
- Start the new worker processes with a new configuration
- Gracefully shutdown the old worker processes
"""
self.log.info("Hang up: %s", self.master_name)
self.reload()
def handle_quit(self):
"SIGQUIT handling"
raise StopIteration
def handle_int(self):
"SIGINT handling"
self.stop(False)
raise StopIteration
def handle_term(self):
"SIGTERM handling"
self.stop(False)
raise StopIteration
def handle_ttin(self):
"""\
SIGTTIN handling.
Increases the number of workers by one.
"""
self.num_workers += 1
self.manage_workers()
def handle_ttou(self):
"""\
SIGTTOU handling.
Decreases the number of workers by one.
"""
if self.num_workers <= 1:
return
self.num_workers -= 1
self.manage_workers()
def handle_usr1(self):
"""\
SIGUSR1 handling.
Kill all workers by sending them a SIGUSR1
"""
self.kill_workers(signal.SIGUSR1)
self.log.reopen_files()
def handle_usr2(self):
"""\
SIGUSR2 handling.
Creates a new master/worker set as a slave of the current
master without affecting old workers. Use this to do live
deployment with the ability to backout a change.
"""
self.reexec()
def handle_winch(self):
"SIGWINCH handling"
if os.getppid() == 1 or os.getpgrp() != os.getpid():
self.log.info("graceful stop of workers")
self.num_workers = 0
self.kill_workers(signal.SIGQUIT)
else:
self.log.info("SIGWINCH ignored. Not daemonized")
def wakeup(self):
"""\
Wake up the arbiter by writing to the PIPE
"""
try:
os.write(self.PIPE[1], '.')
except IOError, e:
if e.errno not in [errno.EAGAIN, errno.EINTR]:
raise
def halt(self, reason=None, exit_status=0):
""" halt arbiter """
self.stop()
self.log.info("Shutting down: %s", self.master_name)
if reason is not None:
self.log.info("Reason: %s", reason)
if self.pidfile is not None:
self.pidfile.unlink()
sys.exit(exit_status)
def sleep(self):
"""\
Sleep until PIPE is readable or we timeout.
A readable PIPE means a signal occurred.
"""
try:
ready = select.select([self.PIPE[0]], [], [], 1.0)
if not ready[0]:
return
while os.read(self.PIPE[0], 1):
pass
except select.error, e:
if e[0] not in [errno.EAGAIN, errno.EINTR]:
raise
except OSError, e:
if e.errno not in [errno.EAGAIN, errno.EINTR]:
raise
except KeyboardInterrupt:
sys.exit()
def stop(self, graceful=True):
"""\
Stop workers
:attr graceful: boolean, If True (the default) workers will be
killed gracefully (ie. trying to wait for the current connection)
"""
try:
self.LISTENER.close()
except Exception:
pass
self.LISTENER = None
sig = signal.SIGQUIT
if not graceful:
sig = signal.SIGTERM
limit = time.time() + self.cfg.graceful_timeout
while self.WORKERS and time.time() < limit:
self.kill_workers(sig)
time.sleep(0.1)
self.reap_workers()
self.kill_workers(signal.SIGKILL)
def reexec(self):
"""\
Relaunch the master and workers.
"""
if self.pidfile is not None:
self.pidfile.rename("%s.oldbin" % self.pidfile.fname)
self.reexec_pid = os.fork()
if self.reexec_pid != 0:
self.master_name = "Old Master"
return
os.environ['GUNICORN_FD'] = str(self.LISTENER.fileno())
os.chdir(self.START_CTX['cwd'])
self.cfg.pre_exec(self)
util.closerange(3, self.LISTENER.fileno())
util.closerange(self.LISTENER.fileno()+1, util.get_maxfd())
os.execvpe(self.START_CTX[0], self.START_CTX['args'], os.environ)
def reload(self):
old_address = self.cfg.address
# reload conf
self.app.reload()
self.setup(self.app)
# reopen log files
self.log.reopen_files()
# do we need to change listener ?
if old_address != self.cfg.address:
self.LISTENER.close()
self.LISTENER = create_socket(self.cfg, self.log)
self.log.info("Listening at: %s", self.LISTENER)
# do some actions on reload
self.cfg.on_reload(self)
# unlink pidfile
if self.pidfile is not None:
self.pidfile.unlink()
# create new pidfile
if self.cfg.pidfile is not None:
self.pidfile = Pidfile(self.cfg.pidfile)
self.pidfile.create(self.pid)
# set new proc_name
util._setproctitle("master [%s]" % self.proc_name)
# spawn new workers
for i in range(self.cfg.workers):
self.spawn_worker()
# manage workers
self.manage_workers()
def murder_workers(self):
"""\
Kill unused/idle workers
"""
if not self.timeout:
return
for (pid, worker) in self.WORKERS.items():
try:
if time.time() - worker.tmp.last_update() <= self.timeout:
continue
except ValueError:
continue
self.log.critical("WORKER TIMEOUT (pid:%s)", pid)
self.kill_worker(pid, signal.SIGKILL)
def reap_workers(self):
"""\
Reap workers to avoid zombie processes
"""
try:
while True:
wpid, status = os.waitpid(-1, os.WNOHANG)
if not wpid:
break
if self.reexec_pid == wpid:
self.reexec_pid = 0
else:
# A worker said it cannot boot. We'll shutdown
# to avoid infinite start/stop cycles.
exitcode = status >> 8
if exitcode == self.WORKER_BOOT_ERROR:
reason = "Worker failed to boot."
raise HaltServer(reason, self.WORKER_BOOT_ERROR)
worker = self.WORKERS.pop(wpid, None)
if not worker:
continue
worker.tmp.close()
except OSError, e:
if e.errno == errno.ECHILD:
pass
def manage_workers(self):
"""\
Maintain the number of workers by spawning or killing
as required.
"""
if len(self.WORKERS.keys()) < self.num_workers:
self.spawn_workers()
workers = self.WORKERS.items()
workers.sort(key=lambda w: w[1].age)
while len(workers) > self.num_workers:
(pid, _) = workers.pop(0)
self.kill_worker(pid, signal.SIGQUIT)
def spawn_worker(self):
self.worker_age += 1
worker = self.worker_class(self.worker_age, self.pid, self.LISTENER,
self.app, self.timeout/2.0,
self.cfg, self.log)
self.cfg.pre_fork(self, worker)
pid = os.fork()
if pid != 0:
self.WORKERS[pid] = worker
return pid
# Process Child
worker_pid = os.getpid()
try:
util._setproctitle("worker [%s]" % self.proc_name)
self.log.info("Booting worker with pid: %s", worker_pid)
self.cfg.post_fork(self, worker)
worker.init_process()
sys.exit(0)
except SystemExit:
raise
except:
self.log.exception("Exception in worker process:\n%s",
traceback.format_exc())
if not worker.booted:
sys.exit(self.WORKER_BOOT_ERROR)
sys.exit(-1)
finally:
self.log.info("Worker exiting (pid: %s)", worker_pid)
try:
worker.tmp.close()
self.cfg.worker_exit(self, worker)
except:
pass
def spawn_workers(self):
"""\
Spawn new workers as needed.
This is where a worker process leaves the main loop
of the master process.
"""
for i in range(self.num_workers - len(self.WORKERS.keys())):
self.spawn_worker()
def kill_workers(self, sig):
"""\
Kill all workers with the signal `sig`
:attr sig: `signal.SIG*` value
"""
for pid in self.WORKERS.keys():
self.kill_worker(pid, sig)
def kill_worker(self, pid, sig):
"""\
Kill a worker
:attr pid: int, worker pid
:attr sig: `signal.SIG*` value
"""
try:
os.kill(pid, sig)
except OSError, e:
if e.errno == errno.ESRCH:
try:
worker = self.WORKERS.pop(pid)
worker.tmp.close()
self.cfg.worker_exit(self, worker)
return
except (KeyError, OSError):
return
raise