mirror of
/repos/baseimage-docker.git
synced 2026-02-27 17:41:59 +01:00
Improve the init system: support skipping startup files and running a custom main command
This commit is contained in:
Hongli Lai (Phusion)
279
image/my_init
279
image/my_init
@@ -1,8 +1,42 @@
|
||||
#!/usr/bin/python2
|
||||
import os, sys, stat, signal, errno
|
||||
import os, sys, stat, signal, errno, argparse, time
|
||||
|
||||
pid = None
|
||||
status = None
|
||||
KILL_PROCESS_TIMEOUT = 5
|
||||
KILL_ALL_PROCESSES_TIMEOUT = 5
|
||||
|
||||
LOG_LEVEL_ERROR = 1
|
||||
LOG_LEVEL_WARN = 1
|
||||
LOG_LEVEL_INFO = 2
|
||||
LOG_LEVEL_DEBUG = 3
|
||||
|
||||
log_level = None
|
||||
|
||||
class AlarmException(Exception):
|
||||
pass
|
||||
|
||||
def error(message):
|
||||
if log_level >= LOG_LEVEL_ERROR:
|
||||
sys.stderr.write("*** %s\n" % message)
|
||||
|
||||
def warn(message):
|
||||
if log_level >= LOG_LEVEL_WARN:
|
||||
print("*** %s" % message)
|
||||
|
||||
def info(message):
|
||||
if log_level >= LOG_LEVEL_INFO:
|
||||
print("*** %s" % message)
|
||||
|
||||
def debug(message):
|
||||
if log_level >= LOG_LEVEL_DEBUG:
|
||||
print("*** %s" % message)
|
||||
|
||||
def ignore_signals_and_raise_keyboard_interrupt(signame):
|
||||
signal.signal(signal.SIGTERM, signal.SIG_IGN)
|
||||
signal.signal(signal.SIGINT, signal.SIG_IGN)
|
||||
raise KeyboardInterrupt(signame)
|
||||
|
||||
def raise_alarm_exception():
|
||||
raise AlarmException('Alarm')
|
||||
|
||||
def listdir(path):
|
||||
try:
|
||||
@@ -20,100 +54,185 @@ def is_exe(path):
|
||||
except OSError:
|
||||
return False
|
||||
|
||||
def reap_child(signum, frame):
|
||||
global pid, status, waiting_for_runit
|
||||
try:
|
||||
result = os.wait3(os.WNOHANG)
|
||||
if result is not None and pid == result[0]:
|
||||
status = result[1]
|
||||
except OSError:
|
||||
pass
|
||||
def waitpid_reap_other_children(pid):
|
||||
done = False
|
||||
status = None
|
||||
while not done:
|
||||
this_pid, status = os.waitpid(-1, 0)
|
||||
done = this_pid == pid
|
||||
return status
|
||||
|
||||
def stop_child_process(name, pid):
|
||||
print("*** Shutting down %s (PID %d)..." % (name, pid))
|
||||
def stop_child_process(name, pid, signo = signal.SIGTERM, time_limit = KILL_PROCESS_TIMEOUT):
|
||||
info("Shutting down %s (PID %d)..." % (name, pid))
|
||||
try:
|
||||
os.kill(pid, signal.SIGHUP)
|
||||
os.kill(pid, signo)
|
||||
except OSError:
|
||||
pass
|
||||
signal.alarm(time_limit)
|
||||
try:
|
||||
try:
|
||||
waitpid_reap_other_children(pid)
|
||||
except OSError:
|
||||
pass
|
||||
except AlarmException:
|
||||
warn("%s (PID %d) did not shut down in time. Forcing it to exit.")
|
||||
try:
|
||||
os.kill(pid, signal.SIGKILL)
|
||||
except OSError:
|
||||
pass
|
||||
try:
|
||||
waitpid_reap_other_children(pid)
|
||||
except OSError:
|
||||
pass
|
||||
finally:
|
||||
signal.alarm(0)
|
||||
|
||||
def run_command_killable(*argv):
|
||||
global pid
|
||||
filename = argv[0]
|
||||
status = None
|
||||
pid = os.spawnvp(os.P_NOWAIT, filename, argv)
|
||||
signal.signal(signal.SIGINT, lambda signum, frame: stop_child_process(filename, pid))
|
||||
signal.signal(signal.SIGTERM, lambda signum, frame: stop_child_process(filename, pid))
|
||||
try:
|
||||
status = waitpid_reap_other_children(pid)
|
||||
except BaseException as s:
|
||||
warn("An error occurred. Aborting.")
|
||||
stop_child_process(filename, pid)
|
||||
raise
|
||||
if status != 0:
|
||||
error("%s failed with exit code %d\n" % (filename, status))
|
||||
sys.exit(1)
|
||||
|
||||
def kill_all_processes(time_limit):
|
||||
info("Killing all processes...")
|
||||
try:
|
||||
os.kill(-1, signal.SIGTERM)
|
||||
except OSError:
|
||||
pass
|
||||
signal.alarm(time_limit)
|
||||
try:
|
||||
# Wait until no more child processes exist.
|
||||
done = False
|
||||
while not done:
|
||||
try:
|
||||
this_pid, status = os.waitpid(-1, 0)
|
||||
done = this_pid == pid
|
||||
os.waitpid(-1, 0)
|
||||
except OSError as e:
|
||||
if e.errno == errno.EINTR:
|
||||
sys.exit(2)
|
||||
if e.errno == errno.ECHILD:
|
||||
done = True
|
||||
else:
|
||||
raise
|
||||
finally:
|
||||
signal.signal(signal.SIGINT, signal.SIG_DFL)
|
||||
signal.signal(signal.SIGTERM, signal.SIG_DFL)
|
||||
if status != 0:
|
||||
sys.stderr.write("*** %s failed with exit code %d\n" % (filename, status))
|
||||
sys.exit(1)
|
||||
|
||||
# Run /etc/my_init.d/*
|
||||
for name in listdir("/etc/my_init.d"):
|
||||
filename = "/etc/my_init.d/" + name
|
||||
if is_exe(filename):
|
||||
print("*** Running %s..." % filename)
|
||||
run_command_killable(filename)
|
||||
|
||||
# Run /etc/rc.local.
|
||||
if is_exe("/etc/rc.local"):
|
||||
print("*** Running /etc/rc.local...")
|
||||
run_command_killable("/etc/rc.local")
|
||||
|
||||
# Start runit.
|
||||
signal.signal(signal.SIGCHLD, reap_child)
|
||||
print("*** Booting runit...")
|
||||
pid = os.spawnl(os.P_NOWAIT, "/usr/bin/runsvdir", "/usr/bin/runsvdir", "-P", "/etc/service", "log: %s" % ('.' * 395))
|
||||
print("*** Runit started as PID %d" % pid)
|
||||
signal.signal(signal.SIGTERM, lambda signum, frame: stop_child_process("runit", pid))
|
||||
|
||||
# Wait for runit, and while waiting, reap any adopted orphans.
|
||||
done = False
|
||||
while not done:
|
||||
try:
|
||||
this_pid, status = os.waitpid(pid, 0)
|
||||
done = True
|
||||
except OSError as e:
|
||||
if e.errno == errno.EINTR:
|
||||
# Try again
|
||||
except AlarmException:
|
||||
warn("Not all processes have exited in time. Forcing them to exit.")
|
||||
try:
|
||||
os.kill(-1, signal.SIGKILL)
|
||||
except OSError:
|
||||
pass
|
||||
finally:
|
||||
signal.alarm(0)
|
||||
|
||||
def run_startup_files():
|
||||
# Run /etc/my_init.d/*
|
||||
for name in listdir("/etc/my_init.d"):
|
||||
filename = "/etc/my_init.d/" + name
|
||||
if is_exe(filename):
|
||||
info("Running %s..." % filename)
|
||||
run_command_killable(filename)
|
||||
|
||||
# Run /etc/rc.local.
|
||||
if is_exe("/etc/rc.local"):
|
||||
info("Running /etc/rc.local...")
|
||||
run_command_killable("/etc/rc.local")
|
||||
|
||||
def start_runit():
|
||||
info("Booting runit daemon...")
|
||||
pid = os.spawnl(os.P_NOWAIT, "/usr/bin/runsvdir", "/usr/bin/runsvdir",
|
||||
"-P", "/etc/service", "log: %s" % ('.' * 395))
|
||||
info("Runit started as PID %d" % pid)
|
||||
return pid
|
||||
|
||||
def wait_for_runit_or_interrupt(pid):
|
||||
try:
|
||||
status = waitpid_reap_other_children(pid)
|
||||
return (True, status)
|
||||
except KeyboardInterrupt:
|
||||
return (False, None)
|
||||
|
||||
def shutdown_runit_services():
|
||||
debug("Begin shutting down runit services...")
|
||||
os.system("/usr/bin/sv down /etc/service/*")
|
||||
|
||||
def wait_for_runit_services():
|
||||
debug("Waiting for runit services to exit...")
|
||||
done = False
|
||||
while not done:
|
||||
done = os.system("/usr/bin/sv status /etc/service/* | grep -q '^run:'") != 0
|
||||
if not done:
|
||||
time.sleep(0.1)
|
||||
|
||||
def main(args):
|
||||
if not args.skip_startup_files:
|
||||
run_startup_files()
|
||||
|
||||
runit_exited = False
|
||||
exit_code = None
|
||||
|
||||
if not args.skip_runit:
|
||||
runit_pid = start_runit()
|
||||
try:
|
||||
if len(args.main_command) == 0:
|
||||
runit_exited, exit_code = wait_for_runit_or_interrupt(runit_pid)
|
||||
if runit_exited:
|
||||
info("Runit exited with code %d" % exit_code)
|
||||
else:
|
||||
# The SIGCHLD handler probably caught it.
|
||||
done = True
|
||||
info("Running %s..." % " ".join(args.main_command))
|
||||
pid = os.spawnvp(os.P_NOWAIT, args.main_command[0], args.main_command)
|
||||
try:
|
||||
exit_code = waitpid_reap_other_children(pid)
|
||||
info("%s exited with exit code %d." % (args.main_command[0], exit_code))
|
||||
except KeyboardInterrupt:
|
||||
stop_child_process(args.main_command[0], pid)
|
||||
except BaseException as s:
|
||||
warn("An error occurred. Aborting.")
|
||||
stop_child_process(args.main_command[0], pid)
|
||||
raise
|
||||
sys.exit(exit_code)
|
||||
finally:
|
||||
if not args.skip_runit:
|
||||
shutdown_runit_services()
|
||||
if not runit_exited:
|
||||
stop_child_process("runit daemon", runit_pid)
|
||||
wait_for_runit_services()
|
||||
|
||||
# Runit has exited. Reset signal handlers.
|
||||
print("*** Runit exited with code %s. Waiting for all services to shut down..." % status)
|
||||
signal.signal(signal.SIGCHLD, signal.SIG_DFL)
|
||||
signal.signal(signal.SIGTERM, signal.SIG_DFL)
|
||||
signal.siginterrupt(signal.SIGCHLD, False)
|
||||
signal.siginterrupt(signal.SIGTERM, False)
|
||||
# Parse options.
|
||||
parser = argparse.ArgumentParser(description = 'Initialize the system.')
|
||||
parser.add_argument('main_command', metavar = 'MAIN_COMMAND', type = str, nargs = '*',
|
||||
help = 'The main command to run. (default: runit)')
|
||||
parser.add_argument('--skip-startup-files', dest = 'skip_startup_files',
|
||||
action = 'store_const', const = True, default = False,
|
||||
help = 'Skip running /etc/my_init.d/* and /etc/rc.local')
|
||||
parser.add_argument('--skip-runit', dest = 'skip_runit',
|
||||
action = 'store_const', const = True, default = False,
|
||||
help = 'Do not run runit services')
|
||||
parser.add_argument('--no-kill-all-on-exit', dest = 'kill_all_on_exit',
|
||||
action = 'store_const', const = False, default = True,
|
||||
help = 'Don\'t kill all processes on the system upon exiting')
|
||||
parser.add_argument('--quiet', dest = 'log_level',
|
||||
action = 'store_const', const = LOG_LEVEL_WARN, default = LOG_LEVEL_INFO,
|
||||
help = 'Only print warnings and errors')
|
||||
args = parser.parse_args()
|
||||
log_level = args.log_level
|
||||
|
||||
# Wait at most 5 seconds for services to shut down.
|
||||
import time
|
||||
|
||||
def shutdown(signum = None, frame = None):
|
||||
global status
|
||||
if status is not None:
|
||||
sys.exit(status)
|
||||
|
||||
signal.signal(signal.SIGALRM, shutdown)
|
||||
signal.alarm(5)
|
||||
done = False
|
||||
while not done:
|
||||
done = os.system("/usr/bin/sv status /etc/service/* | grep -q '^run:'") != 0
|
||||
if not done:
|
||||
time.sleep(0.5)
|
||||
shutdown()
|
||||
if args.skip_runit and len(args.main_command) == 0:
|
||||
error("When --skip-runit is given, you must also pass a main command.")
|
||||
sys.exit(1)
|
||||
|
||||
# Run main function.
|
||||
signal.signal(signal.SIGTERM, lambda signum, frame: ignore_signals_and_raise_keyboard_interrupt('SIGTERM'))
|
||||
signal.signal(signal.SIGINT, lambda signum, frame: ignore_signals_and_raise_keyboard_interrupt('SIGINT'))
|
||||
signal.signal(signal.SIGALRM, lambda signum, frame: raise_alarm_exception())
|
||||
try:
|
||||
main(args)
|
||||
except KeyboardInterrupt:
|
||||
warn("Init system aborted.")
|
||||
exit(2)
|
||||
finally:
|
||||
if args.kill_all_on_exit:
|
||||
kill_all_processes(KILL_ALL_PROCESSES_TIMEOUT)
|
||||
|
||||
Reference in New Issue
Block a user