def maybe_restart_mirroring_script() -> None:
if os.stat(os.path.join(
options.stamp_path, "stamps",
"restart_stamp")).st_mtime > start_time or (
(options.user == "tabbott" or options.user == "tabbott/extra")
and os.stat(
os.path.join(options.stamp_path, "stamps",
"tabbott_stamp")).st_mtime > start_time):
logger.warning("")
logger.warning(
"zephyr mirroring script has been updated; restarting...")
maybe_kill_child()
try:
zephyr._z.cancelSubs()
except OSError:
# We don't care whether we failed to cancel subs properly, but we should log it
logger.exception("")
backoff = RandomExponentialBackoff(maximum_retries=3, )
while backoff.keep_going():
try:
os.execvp(os.path.abspath(__file__), sys.argv)
# No need for backoff.succeed, since this can't be reached
except Exception:
logger.exception(
"Error restarting mirroring script; trying again... Traceback:"
)
backoff.fail()
raise Exception("Failed to reload too many times, aborting!")
def zulip_to_zephyr(options: int) -> NoReturn:
# Sync messages from zulip to zephyr
logger.info("Starting syncing messages.")
backoff = RandomExponentialBackoff(timeout_success_equivalent=120)
while True:
try:
zulip_client.call_on_each_message(maybe_forward_to_zephyr)
except Exception:
logger.exception("Error syncing messages:")
backoff.fail()
def process_loop(log: Optional[IO[Any]]) -> NoReturn:
restart_check_count = 0
last_check_time = time.time()
recieve_backoff = RandomExponentialBackoff()
while True:
select.select([zephyr._z.getFD()], [], [], 15)
try:
process_backoff = RandomExponentialBackoff()
# Fetch notices from the queue until its empty
while True:
notice = zephyr.receive(block=False)
recieve_backoff.succeed()
if notice is None:
break
try:
process_notice(notice, log)
process_backoff.succeed()
except Exception:
logger.exception("Error relaying zephyr:")
process_backoff.fail()
except Exception:
logger.exception("Error checking for new zephyrs:")
recieve_backoff.fail()
continue
if time.time() - last_check_time > 15:
last_check_time = time.time()
try:
maybe_restart_mirroring_script()
if restart_check_count > 0:
logger.info(
"Stopped getting errors checking whether restart is required."
)
restart_check_count = 0
except Exception:
if restart_check_count < 5:
logger.exception(
"Error checking whether restart is required:")
restart_check_count += 1
if options.forward_class_messages:
try:
update_subscriptions()
except Exception:
logger.exception(
"Error updating subscriptions from Zulip:")
"""We actually want to exit, so run os._exit (so as not to be caught and restarted)"""
os._exit(1)
signal.signal(signal.SIGINT, die)
args = [os.path.join(os.path.dirname(sys.argv[0]), "jabber_mirror_backend.py")]
args.extend(sys.argv[1:])
backoff = RandomExponentialBackoff(timeout_success_equivalent=300)
while backoff.keep_going():
print("Starting Jabber mirroring bot")
try:
ret = subprocess.call(args)
except Exception:
traceback.print_exc()
else:
if ret == 2:
# Don't try again on initial configuration errors
sys.exit(ret)
backoff.fail()
print("")
print("")
print(
"ERROR: The Jabber mirroring bot is unable to continue mirroring Jabber.")
print("Please contact [email protected] if you need assistance.")
print("")
sys.exit(1)
print("Starting parallel zephyr class mirroring bot")
jobs = list("0123456789abcdef")
def run_job(shard):
# type: (str) -> int
subprocess.call(args + ["--shard=%s" % (shard,)])
return 0
for (status, job) in run_parallel(run_job, jobs, threads=16):
print("A mirroring shard died!")
pass
sys.exit(0)
backoff = RandomExponentialBackoff(timeout_success_equivalent=300)
while backoff.keep_going():
print("Starting zephyr mirroring bot")
try:
subprocess.call(args)
except Exception:
traceback.print_exc()
backoff.fail()
error_message = """
ERROR: The Zephyr mirroring bot is unable to continue mirroring Zephyrs.
This is often caused by failing to maintain unexpired Kerberos tickets
or AFS tokens. See https://zulipchat.com/zephyr for documentation on how to
maintain unexpired Kerberos tickets and AFS tokens.
"""
print(error_message)
sys.exit(1)