ominfra 0.0.0.dev75__py3-none-any.whl → 0.0.0.dev77__py3-none-any.whl

ominfra/supervisor/_supervisor.py

@@ -0,0 +1,3305 @@
+#!/usr/bin/env python3
+# noinspection DuplicatedCode
+# @omlish-lite
+# @omlish-script
+# @omlish-amalg-output supervisor.py
+# ruff: noqa: N802 UP006 UP007 UP036
+import abc
+import contextlib
+import dataclasses as dc
+import datetime
+import errno
+import fcntl
+import functools
+import grp
+import json
+import logging
+import os
+import pwd
+import re
+import resource
+import select
+import shlex
+import signal
+import stat
+import sys
+import tempfile
+import threading
+import time
+import traceback
+import types
+import typing as ta
+import warnings
+
+
+########################################
+
+
+if sys.version_info < (3, 8):
+    raise OSError(
+        f'Requires python (3, 8), got {sys.version_info} from {sys.executable}')  # noqa
+
+
+########################################
+
+
+# ../compat.py
+T = ta.TypeVar('T')
+
+# ../states.py
+ProcessState = int  # ta.TypeAlias
+SupervisorState = int  # ta.TypeAlias
+
+
+########################################
+# ../compat.py
+
+
+def as_bytes(s: ta.Union[str, bytes], encoding: str = 'utf8') -> bytes:
+    if isinstance(s, bytes):
+        return s
+    else:
+        return s.encode(encoding)
+
+
+def as_string(s: ta.Union[str, bytes], encoding='utf8') -> str:
+    if isinstance(s, str):
+        return s
+    else:
+        return s.decode(encoding)
+
+
+def compact_traceback() -> ta.Tuple[
+    ta.Tuple[str, str, int],
+    ta.Type[BaseException],
+    BaseException,
+    types.TracebackType,
+]:
+    t, v, tb = sys.exc_info()
+    tbinfo = []
+    if not tb:
+        raise RuntimeError('No traceback')
+    while tb:
+        tbinfo.append((
+            tb.tb_frame.f_code.co_filename,
+            tb.tb_frame.f_code.co_name,
+            str(tb.tb_lineno),
+        ))
+        tb = tb.tb_next
+
+    # just to be safe
+    del tb
+
+    file, function, line = tbinfo[-1]
+    info = ' '.join(['[%s|%s|%s]' % x for x in tbinfo])  # noqa
+    return (file, function, line), t, v, info  # type: ignore
+
+
+def find_prefix_at_end(haystack: bytes, needle: bytes) -> int:
+    l = len(needle) - 1
+    while l and not haystack.endswith(needle[:l]):
+        l -= 1
+    return l
+
+
+class ExitNow(Exception):  # noqa
+    pass
+
+
+##
+
+
+def decode_wait_status(sts: int) -> ta.Tuple[int, str]:
+    """
+    Decode the status returned by wait() or waitpid().
+
+    Return a tuple (exitstatus, message) where exitstatus is the exit status, or -1 if the process was killed by a
+    signal; and message is a message telling what happened.  It is the caller's responsibility to display the message.
+    """
+    if os.WIFEXITED(sts):
+        es = os.WEXITSTATUS(sts) & 0xffff
+        msg = f'exit status {es}'
+        return es, msg
+    elif os.WIFSIGNALED(sts):
+        sig = os.WTERMSIG(sts)
+        msg = f'terminated by {signame(sig)}'
+        if hasattr(os, 'WCOREDUMP'):
+            iscore = os.WCOREDUMP(sts)
+        else:
+            iscore = bool(sts & 0x80)
+        if iscore:
+            msg += ' (core dumped)'
+        return -1, msg
+    else:
+        msg = 'unknown termination cause 0x%04x' % sts  # noqa
+        return -1, msg
+
+
+_signames: ta.Optional[ta.Mapping[int, str]] = None
+
+
+def signame(sig: int) -> str:
+    global _signames
+    if _signames is None:
+        _signames = _init_signames()
+    return _signames.get(sig) or 'signal %d' % sig
+
+
+def _init_signames() -> ta.Dict[int, str]:
+    d = {}
+    for k, v in signal.__dict__.items():
+        k_startswith = getattr(k, 'startswith', None)
+        if k_startswith is None:
+            continue
+        if k_startswith('SIG') and not k_startswith('SIG_'):
+            d[v] = k
+    return d
+
+
+class SignalReceiver:
+    def __init__(self) -> None:
+        super().__init__()
+        self._signals_recvd: ta.List[int] = []
+
+    def receive(self, sig: int, frame: ta.Any) -> None:
+        if sig not in self._signals_recvd:
+            self._signals_recvd.append(sig)
+
+    def install(self, *sigs: int) -> None:
+        for sig in sigs:
+            signal.signal(sig, self.receive)
+
+    def get_signal(self) -> ta.Optional[int]:
+        if self._signals_recvd:
+            sig = self._signals_recvd.pop(0)
+        else:
+            sig = None
+        return sig
+
+
+def readfd(fd: int) -> bytes:
+    try:
+        data = os.read(fd, 2 << 16)  # 128K
+    except OSError as why:
+        if why.args[0] not in (errno.EWOULDBLOCK, errno.EBADF, errno.EINTR):
+            raise
+        data = b''
+    return data
+
+
+def try_unlink(path: str) -> bool:
+    try:
+        os.unlink(path)
+    except OSError:
+        return False
+    return True
+
+
+def close_fd(fd: int) -> bool:
+    try:
+        os.close(fd)
+    except OSError:
+        return False
+    return True
+
+
+def mktempfile(suffix: str, prefix: str, dir: str) -> str:  # noqa
+    fd, filename = tempfile.mkstemp(suffix, prefix, dir)
+    os.close(fd)
+    return filename
+
+
+def real_exit(code: int) -> None:
+    os._exit(code)  # noqa
+
+
+def get_path() -> ta.Sequence[str]:
+    """Return a list corresponding to $PATH, or a default."""
+    path = ['/bin', '/usr/bin', '/usr/local/bin']
+    if 'PATH' in os.environ:
+        p = os.environ['PATH']
+        if p:
+            path = p.split(os.pathsep)
+    return path
+
+
+def normalize_path(v: str) -> str:
+    return os.path.normpath(os.path.abspath(os.path.expanduser(v)))
+
+
+ANSI_ESCAPE_BEGIN = b'\x1b['
+ANSI_TERMINATORS = (b'H', b'f', b'A', b'B', b'C', b'D', b'R', b's', b'u', b'J', b'K', b'h', b'l', b'p', b'm')
+
+
+def strip_escapes(s):
+    """Remove all ANSI color escapes from the given string."""
+    result = b''
+    show = 1
+    i = 0
+    l = len(s)
+    while i < l:
+        if show == 0 and s[i:i + 1] in ANSI_TERMINATORS:
+            show = 1
+        elif show:
+            n = s.find(ANSI_ESCAPE_BEGIN, i)
+            if n == -1:
+                return result + s[i:]
+            else:
+                result = result + s[i:n]
+                i = n
+                show = 0
+        i += 1
+    return result
+
+
+########################################
+# ../datatypes.py
+
+
+class Automatic:
+    pass
+
+
+class Syslog:
+    """TODO deprecated; remove this special 'syslog' filename in the future"""
+
+
+LOGFILE_NONES = ('none', 'off', None)
+LOGFILE_AUTOS = (Automatic, 'auto')
+LOGFILE_SYSLOGS = (Syslog, 'syslog')
+
+
+def logfile_name(val):
+    if hasattr(val, 'lower'):
+        coerced = val.lower()
+    else:
+        coerced = val
+
+    if coerced in LOGFILE_NONES:
+        return None
+    elif coerced in LOGFILE_AUTOS:
+        return Automatic
+    elif coerced in LOGFILE_SYSLOGS:
+        return Syslog
+    else:
+        return existing_dirpath(val)
+
+
+def name_to_uid(name: str) -> int:
+    try:
+        uid = int(name)
+    except ValueError:
+        try:
+            pwdrec = pwd.getpwnam(name)
+        except KeyError:
+            raise ValueError(f'Invalid user name {name}')  # noqa
+        uid = pwdrec[2]
+    else:
+        try:
+            pwd.getpwuid(uid)  # check if uid is valid
+        except KeyError:
+            raise ValueError(f'Invalid user id {name}')  # noqa
+    return uid
+
+
+def name_to_gid(name: str) -> int:
+    try:
+        gid = int(name)
+    except ValueError:
+        try:
+            grprec = grp.getgrnam(name)
+        except KeyError:
+            raise ValueError(f'Invalid group name {name}')  # noqa
+        gid = grprec[2]
+    else:
+        try:
+            grp.getgrgid(gid)  # check if gid is valid
+        except KeyError:
+            raise ValueError(f'Invalid group id {name}')  # noqa
+    return gid
+
+
+def gid_for_uid(uid: int) -> int:
+    pwrec = pwd.getpwuid(uid)
+    return pwrec[3]
+
+
+def octal_type(arg: ta.Union[str, int]) -> int:
+    if isinstance(arg, int):
+        return arg
+    try:
+        return int(arg, 8)
+    except (TypeError, ValueError):
+        raise ValueError(f'{arg} can not be converted to an octal type')  # noqa
+
+
+def existing_directory(v: str) -> str:
+    nv = os.path.expanduser(v)
+    if os.path.isdir(nv):
+        return nv
+    raise ValueError(f'{v} is not an existing directory')
+
+
+def existing_dirpath(v: str) -> str:
+    nv = os.path.expanduser(v)
+    dir = os.path.dirname(nv)  # noqa
+    if not dir:
+        # relative pathname with no directory component
+        return nv
+    if os.path.isdir(dir):
+        return nv
+    raise ValueError(f'The directory named as part of the path {v} does not exist')
+
+
+def logging_level(value: ta.Union[str, int]) -> int:
+    if isinstance(value, int):
+        return value
+    s = str(value).lower()
+    level = logging.getLevelNamesMapping().get(s.upper())
+    if level is None:
+        raise ValueError(f'bad logging level name {value!r}')
+    return level
+
+
+class SuffixMultiplier:
+    # d is a dictionary of suffixes to integer multipliers.  If no suffixes match, default is the multiplier.  Matches
+    # are case insensitive.  Return values are in the fundamental unit.
+    def __init__(self, d, default=1):
+        super().__init__()
+        self._d = d
+        self._default = default
+        # all keys must be the same size
+        self._keysz = None
+        for k in d:
+            if self._keysz is None:
+                self._keysz = len(k)
+            elif self._keysz != len(k):  # type: ignore
+                raise ValueError(k)
+
+    def __call__(self, v: ta.Union[str, int]) -> int:
+        if isinstance(v, int):
+            return v
+        v = v.lower()
+        for s, m in self._d.items():
+            if v[-self._keysz:] == s:  # type: ignore
+                return int(v[:-self._keysz]) * m  # type: ignore
+        return int(v) * self._default
+
+
+byte_size = SuffixMultiplier({
+    'kb': 1024,
+    'mb': 1024 * 1024,
+    'gb': 1024 * 1024 * 1024,
+})
+
+
+# all valid signal numbers
+SIGNUMS = [getattr(signal, k) for k in dir(signal) if k.startswith('SIG')]
+
+
+def signal_number(value: ta.Union[int, str]) -> int:
+    try:
+        num = int(value)
+    except (ValueError, TypeError):
+        name = value.strip().upper()  # type: ignore
+        if not name.startswith('SIG'):
+            name = f'SIG{name}'
+        num = getattr(signal, name, None)  # type: ignore
+        if num is None:
+            raise ValueError(f'value {value!r} is not a valid signal name')  # noqa
+    if num not in SIGNUMS:
+        raise ValueError(f'value {value!r} is not a valid signal number')
+    return num
+
+
+class RestartWhenExitUnexpected:
+    pass
+
+
+class RestartUnconditionally:
+    pass
+
+
+########################################
+# ../exceptions.py
+
+
+class ProcessError(Exception):
+    """ Specialized exceptions used when attempting to start a process """
+
+
+class BadCommandError(ProcessError):
+    """ Indicates the command could not be parsed properly. """
+
+
+class NotExecutableError(ProcessError):
+    """ Indicates that the filespec cannot be executed because its path
+    resolves to a file which is not executable, or which is a directory. """
+
+
+class NotFoundError(ProcessError):
+    """ Indicates that the filespec cannot be executed because it could not be found """
+
+
+class NoPermissionError(ProcessError):
+    """
+    Indicates that the file cannot be executed because the supervisor process does not possess the appropriate UNIX
+    filesystem permission to execute the file.
+    """
+
+
+########################################
+# ../poller.py
+
+
+log = logging.getLogger(__name__)
+
+
+class BasePoller(abc.ABC):
+
+    def __init__(self) -> None:
+        super().__init__()
+
+    @abc.abstractmethod
+    def register_readable(self, fd: int) -> None:
+        raise NotImplementedError
+
+    @abc.abstractmethod
+    def register_writable(self, fd: int) -> None:
+        raise NotImplementedError
+
+    @abc.abstractmethod
+    def unregister_readable(self, fd: int) -> None:
+        raise NotImplementedError
+
+    @abc.abstractmethod
+    def unregister_writable(self, fd: int) -> None:
+        raise NotImplementedError
+
+    @abc.abstractmethod
+    def poll(self, timeout: ta.Optional[float]) -> ta.Tuple[ta.List[int], ta.List[int]]:
+        raise NotImplementedError
+
+    def before_daemonize(self) -> None:  # noqa
+        pass
+
+    def after_daemonize(self) -> None:  # noqa
+        pass
+
+    def close(self) -> None:  # noqa
+        pass
+
+
+class SelectPoller(BasePoller):
+
+    def __init__(self) -> None:
+        super().__init__()
+
+        self._readables: ta.Set[int] = set()
+        self._writables: ta.Set[int] = set()
+
+    def register_readable(self, fd: int) -> None:
+        self._readables.add(fd)
+
+    def register_writable(self, fd: int) -> None:
+        self._writables.add(fd)
+
+    def unregister_readable(self, fd: int) -> None:
+        self._readables.discard(fd)
+
+    def unregister_writable(self, fd: int) -> None:
+        self._writables.discard(fd)
+
+    def unregister_all(self) -> None:
+        self._readables.clear()
+        self._writables.clear()
+
+    def poll(self, timeout: ta.Optional[float]) -> ta.Tuple[ta.List[int], ta.List[int]]:
+        try:
+            r, w, x = select.select(
+                self._readables,
+                self._writables,
+                [], timeout,
+            )
+        except OSError as err:
+            if err.args[0] == errno.EINTR:
+                log.debug('EINTR encountered in poll')
+                return [], []
+            if err.args[0] == errno.EBADF:
+                log.debug('EBADF encountered in poll')
+                self.unregister_all()
+                return [], []
+            raise
+        return r, w
+
+
+class PollPoller(BasePoller):
+    _READ = select.POLLIN | select.POLLPRI | select.POLLHUP
+    _WRITE = select.POLLOUT
+
+    def __init__(self) -> None:
+        super().__init__()
+
+        self._poller = select.poll()
+        self._readables: set[int] = set()
+        self._writables: set[int] = set()
+
+    def register_readable(self, fd: int) -> None:
+        self._poller.register(fd, self._READ)
+        self._readables.add(fd)
+
+    def register_writable(self, fd: int) -> None:
+        self._poller.register(fd, self._WRITE)
+        self._writables.add(fd)
+
+    def unregister_readable(self, fd: int) -> None:
+        self._readables.discard(fd)
+        self._poller.unregister(fd)
+        if fd in self._writables:
+            self._poller.register(fd, self._WRITE)
+
+    def unregister_writable(self, fd: int) -> None:
+        self._writables.discard(fd)
+        self._poller.unregister(fd)
+        if fd in self._readables:
+            self._poller.register(fd, self._READ)
+
+    def poll(self, timeout: ta.Optional[float]) -> ta.Tuple[ta.List[int], ta.List[int]]:
+        fds = self._poll_fds(timeout)  # type: ignore
+        readables, writables = [], []
+        for fd, eventmask in fds:
+            if self._ignore_invalid(fd, eventmask):
+                continue
+            if eventmask & self._READ:
+                readables.append(fd)
+            if eventmask & self._WRITE:
+                writables.append(fd)
+        return readables, writables
+
+    def _poll_fds(self, timeout: float) -> ta.List[ta.Tuple[int, int]]:
+        try:
+            return self._poller.poll(timeout * 1000)
+        except OSError as err:
+            if err.args[0] == errno.EINTR:
+                log.debug('EINTR encountered in poll')
+                return []
+            raise
+
+    def _ignore_invalid(self, fd: int, eventmask: int) -> bool:
+        if eventmask & select.POLLNVAL:
+            # POLLNVAL means `fd` value is invalid, not open. When a process quits it's `fd`s are closed so there is no
+            # more reason to keep this `fd` registered If the process restarts it's `fd`s are registered again.
+            self._poller.unregister(fd)
+            self._readables.discard(fd)
+            self._writables.discard(fd)
+            return True
+        return False
+
+
+class KqueuePoller(BasePoller):
+    max_events = 1000
+
+    def __init__(self) -> None:
+        super().__init__()
+
+        self._kqueue: ta.Optional[ta.Any] = select.kqueue()
+        self._readables: set[int] = set()
+        self._writables: set[int] = set()
+
+    def register_readable(self, fd: int) -> None:
+        self._readables.add(fd)
+        kevent = select.kevent(fd, filter=select.KQ_FILTER_READ, flags=select.KQ_EV_ADD)
+        self._kqueue_control(fd, kevent)
+
+    def register_writable(self, fd: int) -> None:
+        self._writables.add(fd)
+        kevent = select.kevent(fd, filter=select.KQ_FILTER_WRITE, flags=select.KQ_EV_ADD)
+        self._kqueue_control(fd, kevent)
+
+    def unregister_readable(self, fd: int) -> None:
+        kevent = select.kevent(fd, filter=select.KQ_FILTER_READ, flags=select.KQ_EV_DELETE)
+        self._readables.discard(fd)
+        self._kqueue_control(fd, kevent)
+
+    def unregister_writable(self, fd: int) -> None:
+        kevent = select.kevent(fd, filter=select.KQ_FILTER_WRITE, flags=select.KQ_EV_DELETE)
+        self._writables.discard(fd)
+        self._kqueue_control(fd, kevent)
+
+    def _kqueue_control(self, fd: int, kevent: 'select.kevent') -> None:
+        try:
+            self._kqueue.control([kevent], 0)  # type: ignore
+        except OSError as error:
+            if error.errno == errno.EBADF:
+                log.debug('EBADF encountered in kqueue. Invalid file descriptor %s', fd)
+            else:
+                raise
+
+    def poll(self, timeout: ta.Optional[float]) -> ta.Tuple[ta.List[int], ta.List[int]]:
+        readables, writables = [], []  # type: ignore
+
+        try:
+            kevents = self._kqueue.control(None, self.max_events, timeout)  # type: ignore
+        except OSError as error:
+            if error.errno == errno.EINTR:
+                log.debug('EINTR encountered in poll')
+                return readables, writables
+            raise
+
+        for kevent in kevents:
+            if kevent.filter == select.KQ_FILTER_READ:
+                readables.append(kevent.ident)
+            if kevent.filter == select.KQ_FILTER_WRITE:
+                writables.append(kevent.ident)
+
+        return readables, writables
+
+    def before_daemonize(self) -> None:
+        self.close()
+
+    def after_daemonize(self) -> None:
+        self._kqueue = select.kqueue()
+        for fd in self._readables:
+            self.register_readable(fd)
+        for fd in self._writables:
+            self.register_writable(fd)
+
+    def close(self) -> None:
+        self._kqueue.close()  # type: ignore
+        self._kqueue = None
+
+
+Poller: ta.Type[BasePoller]
+if hasattr(select, 'kqueue'):
+    Poller = KqueuePoller
+elif hasattr(select, 'poll'):
+    Poller = PollPoller
+else:
+    Poller = SelectPoller
+
+
+########################################
+# ../../../omlish/lite/check.py
+
+
+def check_isinstance(v: T, spec: ta.Union[ta.Type[T], tuple]) -> T:
+    if not isinstance(v, spec):
+        raise TypeError(v)
+    return v
+
+
+def check_not_isinstance(v: T, spec: ta.Union[type, tuple]) -> T:
+    if isinstance(v, spec):
+        raise TypeError(v)
+    return v
+
+
+def check_not_none(v: ta.Optional[T]) -> T:
+    if v is None:
+        raise ValueError
+    return v
+
+
+def check_not(v: ta.Any) -> None:
+    if v:
+        raise ValueError(v)
+    return v
+
+
+def check_non_empty_str(v: ta.Optional[str]) -> str:
+    if not v:
+        raise ValueError
+    return v
+
+
+def check_state(v: bool, msg: str = 'Illegal state') -> None:
+    if not v:
+        raise ValueError(msg)
+
+
+########################################
+# ../../../omlish/lite/json.py
+
+
+##
+
+
+JSON_PRETTY_INDENT = 2
+
+JSON_PRETTY_KWARGS: ta.Mapping[str, ta.Any] = dict(
+    indent=JSON_PRETTY_INDENT,
+)
+
+json_dump_pretty: ta.Callable[..., bytes] = functools.partial(json.dump, **JSON_PRETTY_KWARGS)  # type: ignore
+json_dumps_pretty: ta.Callable[..., str] = functools.partial(json.dumps, **JSON_PRETTY_KWARGS)
+
+
+##
+
+
+JSON_COMPACT_SEPARATORS = (',', ':')
+
+JSON_COMPACT_KWARGS: ta.Mapping[str, ta.Any] = dict(
+    indent=None,
+    separators=JSON_COMPACT_SEPARATORS,
+)
+
+json_dump_compact: ta.Callable[..., bytes] = functools.partial(json.dump, **JSON_COMPACT_KWARGS)  # type: ignore
+json_dumps_compact: ta.Callable[..., str] = functools.partial(json.dumps, **JSON_COMPACT_KWARGS)
+
+
+########################################
+# ../configs.py
+
+
+@dc.dataclass(frozen=True)
+class ServerConfig:
+    user: ta.Optional[str] = None
+    nodaemon: bool = False
+    umask: int = 0o22
+    directory: ta.Optional[str] = None
+    logfile: str = 'supervisord.log'
+    logfile_maxbytes: int = 50 * 1024 * 1024
+    logfile_backups: int = 10
+    loglevel: int = logging.INFO
+    pidfile: str = 'supervisord.pid'
+    identifier: str = 'supervisor'
+    child_logdir: str = '/dev/null'
+    minfds: int = 1024
+    minprocs: int = 200
+    nocleanup: bool = False
+    strip_ansi: bool = False
+    silent: bool = False
+
+    groups: ta.Optional[ta.Sequence['ProcessGroupConfig']] = None
+
+    @classmethod
+    def new(
+            cls,
+            umask: ta.Union[int, str] = 0o22,
+            directory: ta.Optional[str] = None,
+            logfile: str = 'supervisord.log',
+            logfile_maxbytes: ta.Union[int, str] = 50 * 1024 * 1024,
+            loglevel: ta.Union[int, str] = logging.INFO,
+            pidfile: str = 'supervisord.pid',
+            child_logdir: ta.Optional[str] = None,
+            **kwargs: ta.Any,
+    ) -> 'ServerConfig':
+        return cls(
+            umask=octal_type(umask),
+            directory=existing_directory(directory) if directory is not None else None,
+            logfile=existing_dirpath(logfile),
+            logfile_maxbytes=byte_size(logfile_maxbytes),
+            loglevel=logging_level(loglevel),
+            pidfile=existing_dirpath(pidfile),
+            child_logdir=child_logdir if child_logdir else tempfile.gettempdir(),
+            **kwargs,
+        )
+
+
+@dc.dataclass(frozen=True)
+class ProcessGroupConfig:
+    name: str
+
+    priority: int = 999
+
+    processes: ta.Optional[ta.Sequence['ProcessConfig']] = None
+
+
+@dc.dataclass(frozen=True)
+class ProcessConfig:
+    name: str
+    command: str
+
+    uid: ta.Optional[int] = None
+    directory: ta.Optional[str] = None
+    umask: ta.Optional[int] = None
+    priority: int = 999
+
+    autostart: bool = True
+    autorestart: str = 'unexpected'
+
+    startsecs: int = 1
+    startretries: int = 3
+
+    numprocs: int = 1
+    numprocs_start: int = 0
+
+    @dc.dataclass(frozen=True)
+    class Log:
+        file: ta.Optional[str] = None
+        capture_maxbytes: ta.Optional[int] = None
+        events_enabled: bool = False
+        syslog: bool = False
+        backups: ta.Optional[int] = None
+        maxbytes: ta.Optional[int] = None
+
+    stdout: Log = Log()
+    stderr: Log = Log()
+
+    stopsignal: int = signal.SIGTERM
+    stopwaitsecs: int = 10
+    stopasgroup: bool = False
+
+    killasgroup: bool = False
+
+    exitcodes: ta.Iterable[int] = (0,)
+
+    redirect_stderr: bool = False
+
+    environment: ta.Optional[ta.Mapping[str, str]] = None
+
+
+########################################
+# ../states.py
+
+
+##
+
+
+def _names_by_code(states: ta.Any) -> ta.Dict[int, str]:
+    d = {}
+    for name in states.__dict__:
+        if not name.startswith('__'):
+            code = getattr(states, name)
+            d[code] = name
+    return d
+
+
+##
+
+
+class ProcessStates:
+    STOPPED = 0
+    STARTING = 10
+    RUNNING = 20
+    BACKOFF = 30
+    STOPPING = 40
+    EXITED = 100
+    FATAL = 200
+    UNKNOWN = 1000
+
+
+STOPPED_STATES = (
+    ProcessStates.STOPPED,
+    ProcessStates.EXITED,
+    ProcessStates.FATAL,
+    ProcessStates.UNKNOWN,
+)
+
+RUNNING_STATES = (
+    ProcessStates.RUNNING,
+    ProcessStates.BACKOFF,
+    ProcessStates.STARTING,
+)
+
+SIGNALLABLE_STATES = (
+    ProcessStates.RUNNING,
+    ProcessStates.STARTING,
+    ProcessStates.STOPPING,
+)
+
+
+_process_states_by_code = _names_by_code(ProcessStates)
+
+
+def get_process_state_description(code: ProcessState) -> str:
+    return check_not_none(_process_states_by_code.get(code))
+
+
+##
+
+
+class SupervisorStates:
+    FATAL = 2
+    RUNNING = 1
+    RESTARTING = 0
+    SHUTDOWN = -1
+
+
+_supervisor_states_by_code = _names_by_code(SupervisorStates)
+
+
+def get_supervisor_state_description(code: SupervisorState) -> str:
+    return check_not_none(_supervisor_states_by_code.get(code))
+
+
+########################################
+# ../../../omlish/lite/logs.py
+"""
+TODO:
+ - translate json keys
+ - debug
+"""
+
+
+log = logging.getLogger(__name__)
+
+
+##
+
+
+class TidLogFilter(logging.Filter):
+
+    def filter(self, record):
+        record.tid = threading.get_native_id()
+        return True
+
+
+##
+
+
+class JsonLogFormatter(logging.Formatter):
+
+    KEYS: ta.Mapping[str, bool] = {
+        'name': False,
+        'msg': False,
+        'args': False,
+        'levelname': False,
+        'levelno': False,
+        'pathname': False,
+        'filename': False,
+        'module': False,
+        'exc_info': True,
+        'exc_text': True,
+        'stack_info': True,
+        'lineno': False,
+        'funcName': False,
+        'created': False,
+        'msecs': False,
+        'relativeCreated': False,
+        'thread': False,
+        'threadName': False,
+        'processName': False,
+        'process': False,
+    }
+
+    def format(self, record: logging.LogRecord) -> str:
+        dct = {
+            k: v
+            for k, o in self.KEYS.items()
+            for v in [getattr(record, k)]
+            if not (o and v is None)
+        }
+        return json_dumps_compact(dct)
+
+
+##
+
+
+STANDARD_LOG_FORMAT_PARTS = [
+    ('asctime', '%(asctime)-15s'),
+    ('process', 'pid=%(process)-6s'),
+    ('thread', 'tid=%(thread)x'),
+    ('levelname', '%(levelname)s'),
+    ('name', '%(name)s'),
+    ('separator', '::'),
+    ('message', '%(message)s'),
+]
+
+
+class StandardLogFormatter(logging.Formatter):
+
+    @staticmethod
+    def build_log_format(parts: ta.Iterable[ta.Tuple[str, str]]) -> str:
+        return ' '.join(v for k, v in parts)
+
+    converter = datetime.datetime.fromtimestamp  # type: ignore
+
+    def formatTime(self, record, datefmt=None):
+        ct = self.converter(record.created)  # type: ignore
+        if datefmt:
+            return ct.strftime(datefmt)  # noqa
+        else:
+            t = ct.strftime("%Y-%m-%d %H:%M:%S")  # noqa
+            return '%s.%03d' % (t, record.msecs)
+
+
+##
+
+
+class ProxyLogFilterer(logging.Filterer):
+    def __init__(self, underlying: logging.Filterer) -> None:  # noqa
+        self._underlying = underlying
+
+    @property
+    def underlying(self) -> logging.Filterer:
+        return self._underlying
+
+    @property
+    def filters(self):
+        return self._underlying.filters
+
+    @filters.setter
+    def filters(self, filters):
+        self._underlying.filters = filters
+
+    def addFilter(self, filter):  # noqa
+        self._underlying.addFilter(filter)
+
+    def removeFilter(self, filter):  # noqa
+        self._underlying.removeFilter(filter)
+
+    def filter(self, record):
+        return self._underlying.filter(record)
+
+
+class ProxyLogHandler(ProxyLogFilterer, logging.Handler):
+    def __init__(self, underlying: logging.Handler) -> None:  # noqa
+        ProxyLogFilterer.__init__(self, underlying)
+
+    _underlying: logging.Handler
+
+    @property
+    def underlying(self) -> logging.Handler:
+        return self._underlying
+
+    def get_name(self):
+        return self._underlying.get_name()
+
+    def set_name(self, name):
+        self._underlying.set_name(name)
+
+    @property
+    def name(self):
+        return self._underlying.name
+
+    @property
+    def level(self):
+        return self._underlying.level
+
+    @level.setter
+    def level(self, level):
+        self._underlying.level = level
+
+    @property
+    def formatter(self):
+        return self._underlying.formatter
+
+    @formatter.setter
+    def formatter(self, formatter):
+        self._underlying.formatter = formatter
+
+    def createLock(self):
+        self._underlying.createLock()
+
+    def acquire(self):
+        self._underlying.acquire()
+
+    def release(self):
+        self._underlying.release()
+
+    def setLevel(self, level):
+        self._underlying.setLevel(level)
+
+    def format(self, record):
+        return self._underlying.format(record)
+
+    def emit(self, record):
+        self._underlying.emit(record)
+
+    def handle(self, record):
+        return self._underlying.handle(record)
+
+    def setFormatter(self, fmt):
+        self._underlying.setFormatter(fmt)
+
+    def flush(self):
+        self._underlying.flush()
+
+    def close(self):
+        self._underlying.close()
+
+    def handleError(self, record):
+        self._underlying.handleError(record)
+
+
+##
+
+
+class StandardLogHandler(ProxyLogHandler):
+    pass
+
+
+##
+
+
+@contextlib.contextmanager
+def _locking_logging_module_lock() -> ta.Iterator[None]:
+    if hasattr(logging, '_acquireLock'):
+        logging._acquireLock()  # noqa
+        try:
+            yield
+        finally:
+            logging._releaseLock()  # type: ignore  # noqa
+
+    elif hasattr(logging, '_lock'):
+        # https://github.com/python/cpython/commit/74723e11109a320e628898817ab449b3dad9ee96
+        with logging._lock:  # noqa
+            yield
+
+    else:
+        raise Exception("Can't find lock in logging module")
+
+
+def configure_standard_logging(
+        level: ta.Union[int, str] = logging.INFO,
+        *,
+        json: bool = False,
+        target: ta.Optional[logging.Logger] = None,
+        force: bool = False,
+) -> ta.Optional[StandardLogHandler]:
+    with _locking_logging_module_lock():
+        if target is None:
+            target = logging.root
+
+        #
+
+        if not force:
+            if any(isinstance(h, StandardLogHandler) for h in list(target.handlers)):
+                return None
+
+        #
+
+        handler = logging.StreamHandler()
+
+        #
+
+        formatter: logging.Formatter
+        if json:
+            formatter = JsonLogFormatter()
+        else:
+            formatter = StandardLogFormatter(StandardLogFormatter.build_log_format(STANDARD_LOG_FORMAT_PARTS))
+        handler.setFormatter(formatter)
+
+        #
+
+        handler.addFilter(TidLogFilter())
+
+        #
+
+        target.addHandler(handler)
+
+        #
+
+        if level is not None:
+            target.setLevel(level)
+
+        #
+
+        return StandardLogHandler(handler)
+
+
+########################################
+# ../events.py
+
+
+class EventCallbacks:
+    def __init__(self) -> None:
+        super().__init__()
+
+        self._callbacks: ta.List[ta.Tuple[type, ta.Callable]] = []
+
+    def subscribe(self, type, callback):  # noqa
+        self._callbacks.append((type, callback))
+
+    def unsubscribe(self, type, callback):  # noqa
+        self._callbacks.remove((type, callback))
+
+    def notify(self, event):
+        for type, callback in self._callbacks:  # noqa
+            if isinstance(event, type):
+                callback(event)
+
+    def clear(self):
+        self._callbacks[:] = []
+
+
+EVENT_CALLBACKS = EventCallbacks()
+
+notify_event = EVENT_CALLBACKS.notify
+clear_events = EVENT_CALLBACKS.clear
+
+
+class Event:
+    """Abstract event type """
+
+
+class ProcessLogEvent(Event):
+    """Abstract"""
+    channel: ta.Optional[str] = None
+
+    def __init__(self, process, pid, data):
+        super().__init__()
+        self.process = process
+        self.pid = pid
+        self.data = data
+
+    def payload(self):
+        groupname = ''
+        if self.process.group is not None:
+            groupname = self.process.group.config.name
+        try:
+            data = as_string(self.data)
+        except UnicodeDecodeError:
+            data = f'Undecodable: {self.data!r}'
+        fmt = as_string('processname:%s groupname:%s pid:%s channel:%s\n%s')
+        result = fmt % (
+            as_string(self.process.config.name),
+            as_string(groupname),
+            self.pid,
+            as_string(self.channel),  # type: ignore
+            data,
+        )
+        return result
+
+
+class ProcessLogStdoutEvent(ProcessLogEvent):
+    channel = 'stdout'
+
+
+class ProcessLogStderrEvent(ProcessLogEvent):
+    channel = 'stderr'
+
+
+class ProcessCommunicationEvent(Event):
+    """ Abstract """
+    # event mode tokens
+    BEGIN_TOKEN = b'<!--XSUPERVISOR:BEGIN-->'
+    END_TOKEN = b'<!--XSUPERVISOR:END-->'
+
+    def __init__(self, process, pid, data):
+        super().__init__()
+        self.process = process
+        self.pid = pid
+        self.data = data
+
+    def payload(self):
+        groupname = ''
+        if self.process.group is not None:
+            groupname = self.process.group.config.name
+        try:
+            data = as_string(self.data)
+        except UnicodeDecodeError:
+            data = f'Undecodable: {self.data!r}'
+        return f'processname:{self.process.config.name} groupname:{groupname} pid:{self.pid}\n{data}'
+
+
+class ProcessCommunicationStdoutEvent(ProcessCommunicationEvent):
+    channel = 'stdout'
+
+
+class ProcessCommunicationStderrEvent(ProcessCommunicationEvent):
+    channel = 'stderr'
+
+
+class RemoteCommunicationEvent(Event):
+    def __init__(self, type, data):  # noqa
+        super().__init__()
+        self.type = type
+        self.data = data
+
+    def payload(self):
+        return f'type:{self.type}\n{self.data}'
+
+
+class SupervisorStateChangeEvent(Event):
+    """ Abstract class """
+
+    def payload(self):
+        return ''
+
+
+class SupervisorRunningEvent(SupervisorStateChangeEvent):
+    pass
+
+
+class SupervisorStoppingEvent(SupervisorStateChangeEvent):
+    pass
+
+
+class EventRejectedEvent:  # purposely does not subclass Event
+    def __init__(self, process, event):
+        super().__init__()
+        self.process = process
+        self.event = event
+
+
+class ProcessStateEvent(Event):
+    """ Abstract class, never raised directly """
+    frm = None
+    to = None
+
+    def __init__(self, process, from_state, expected=True):
+        super().__init__()
+        self.process = process
+        self.from_state = from_state
+        self.expected = expected
+        # we eagerly render these so if the process pid, etc changes beneath
+        # us, we stash the values at the time the event was sent
+        self.extra_values = self.get_extra_values()
+
+    def payload(self):
+        groupname = ''
+        if self.process.group is not None:
+            groupname = self.process.group.config.name
+        l = [
+            ('processname', self.process.config.name),
+            ('groupname', groupname),
+            ('from_state', get_process_state_description(self.from_state)),
+        ]
+        l.extend(self.extra_values)
+        s = ' '.join([f'{name}:{val}' for name, val in l])
+        return s
+
+    def get_extra_values(self):
+        return []
+
+
+class ProcessStateFatalEvent(ProcessStateEvent):
+    pass
+
+
+class ProcessStateUnknownEvent(ProcessStateEvent):
+    pass
+
+
+class ProcessStateStartingOrBackoffEvent(ProcessStateEvent):
+    def get_extra_values(self):
+        return [('tries', int(self.process.backoff))]
+
+
+class ProcessStateBackoffEvent(ProcessStateStartingOrBackoffEvent):
+    pass
+
+
+class ProcessStateStartingEvent(ProcessStateStartingOrBackoffEvent):
+    pass
+
+
+class ProcessStateExitedEvent(ProcessStateEvent):
+    def get_extra_values(self):
+        return [('expected', int(self.expected)), ('pid', self.process.pid)]
+
+
+class ProcessStateRunningEvent(ProcessStateEvent):
+    def get_extra_values(self):
+        return [('pid', self.process.pid)]
+
+
+class ProcessStateStoppingEvent(ProcessStateEvent):
+    def get_extra_values(self):
+        return [('pid', self.process.pid)]
+
+
+class ProcessStateStoppedEvent(ProcessStateEvent):
+    def get_extra_values(self):
+        return [('pid', self.process.pid)]
+
+
+class ProcessGroupEvent(Event):
+    def __init__(self, group):
+        super().__init__()
+        self.group = group
+
+    def payload(self):
+        return f'groupname:{self.group}\n'
+
+
+class ProcessGroupAddedEvent(ProcessGroupEvent):
+    pass
+
+
+class ProcessGroupRemovedEvent(ProcessGroupEvent):
+    pass
+
+
+class TickEvent(Event):
+    """ Abstract """
+
+    def __init__(self, when, supervisord):
+        super().__init__()
+        self.when = when
+        self.supervisord = supervisord
+
+    def payload(self):
+        return f'when:{self.when}'
+
+
+class Tick5Event(TickEvent):
+    period = 5
+
+
+class Tick60Event(TickEvent):
+    period = 60
+
+
+class Tick3600Event(TickEvent):
+    period = 3600
+
+
+TICK_EVENTS = [  # imported elsewhere
+    Tick5Event,
+    Tick60Event,
+    Tick3600Event,
+]
+
+
+class EventTypes:
+    EVENT = Event  # abstract
+
+    PROCESS_STATE = ProcessStateEvent  # abstract
+    PROCESS_STATE_STOPPED = ProcessStateStoppedEvent
+    PROCESS_STATE_EXITED = ProcessStateExitedEvent
+    PROCESS_STATE_STARTING = ProcessStateStartingEvent
+    PROCESS_STATE_STOPPING = ProcessStateStoppingEvent
+    PROCESS_STATE_BACKOFF = ProcessStateBackoffEvent
+    PROCESS_STATE_FATAL = ProcessStateFatalEvent
+    PROCESS_STATE_RUNNING = ProcessStateRunningEvent
+    PROCESS_STATE_UNKNOWN = ProcessStateUnknownEvent
+
+    PROCESS_COMMUNICATION = ProcessCommunicationEvent  # abstract
+    PROCESS_COMMUNICATION_STDOUT = ProcessCommunicationStdoutEvent
+    PROCESS_COMMUNICATION_STDERR = ProcessCommunicationStderrEvent
+
+    PROCESS_LOG = ProcessLogEvent
+    PROCESS_LOG_STDOUT = ProcessLogStdoutEvent
+    PROCESS_LOG_STDERR = ProcessLogStderrEvent
+
+    REMOTE_COMMUNICATION = RemoteCommunicationEvent
+
+    SUPERVISOR_STATE_CHANGE = SupervisorStateChangeEvent  # abstract
+    SUPERVISOR_STATE_CHANGE_RUNNING = SupervisorRunningEvent
+    SUPERVISOR_STATE_CHANGE_STOPPING = SupervisorStoppingEvent
+
+    TICK = TickEvent  # abstract
+    TICK_5 = Tick5Event
+    TICK_60 = Tick60Event
+    TICK_3600 = Tick3600Event
+
+    PROCESS_GROUP = ProcessGroupEvent  # abstract
+    PROCESS_GROUP_ADDED = ProcessGroupAddedEvent
+    PROCESS_GROUP_REMOVED = ProcessGroupRemovedEvent
+
+
+def get_event_name_by_type(requested):
+    for name, typ in EventTypes.__dict__.items():
+        if typ is requested:
+            return name
+    return None
+
+
+def register(name, event):
+    setattr(EventTypes, name, event)
+
+
+########################################
+# ../types.py
+
+
+class AbstractServerContext(abc.ABC):
+    @property
+    @abc.abstractmethod
+    def config(self) -> ServerConfig:
+        raise NotImplementedError
+
+    @property
+    @abc.abstractmethod
+    def state(self) -> SupervisorState:
+        raise NotImplementedError
+
+    @abc.abstractmethod
+    def set_state(self, state: SupervisorState) -> None:
+        raise NotImplementedError
+
+    @property
+    @abc.abstractmethod
+    def pid_history(self) -> ta.Dict[int, 'AbstractSubprocess']:
+        raise NotImplementedError
+
+
+class AbstractSubprocess(abc.ABC):
+    @property
+    @abc.abstractmethod
+    def pid(self) -> int:
+        raise NotImplementedError
+
+    @property
+    @abc.abstractmethod
+    def config(self) -> ProcessConfig:
+        raise NotImplementedError
+
+    @property
+    @abc.abstractmethod
+    def context(self) -> AbstractServerContext:
+        raise NotImplementedError
+
+    @abc.abstractmethod
+    def finish(self, sts: int) -> None:
+        raise NotImplementedError
+
+
+########################################
+# ../context.py
+
+
+log = logging.getLogger(__name__)
+
+
+class ServerContext(AbstractServerContext):
+    first = False
+    test = False
+
+    ##
+
+    def __init__(self, config: ServerConfig) -> None:
+        super().__init__()
+
+        self._config = config
+
+        self._pid_history: ta.Dict[int, AbstractSubprocess] = {}
+        self._state: SupervisorState = SupervisorStates.RUNNING
+
+        self.signal_receiver = SignalReceiver()
+
+        self.poller = Poller()
+
+        if self.config.user is not None:
+            uid = name_to_uid(self.config.user)
+            self.uid = uid
+            self.gid = gid_for_uid(uid)
+        else:
+            self.uid = None
+            self.gid = None
+
+        self.unlink_pidfile = False
+
+    @property
+    def config(self) -> ServerConfig:
+        return self._config
+
+    @property
+    def state(self) -> SupervisorState:
+        return self._state
+
+    def set_state(self, state: SupervisorState) -> None:
+        self._state = state
+
+    @property
+    def pid_history(self) -> ta.Dict[int, AbstractSubprocess]:
+        return self._pid_history
+
+    uid: ta.Optional[int]
+    gid: ta.Optional[int]
+
+    ##
+
+    def set_signals(self) -> None:
+        self.signal_receiver.install(
+            signal.SIGTERM,
+            signal.SIGINT,
+            signal.SIGQUIT,
+            signal.SIGHUP,
+            signal.SIGCHLD,
+            signal.SIGUSR2,
+        )
+
+    def waitpid(self) -> ta.Tuple[ta.Optional[int], ta.Optional[int]]:
+        # Need pthread_sigmask here to avoid concurrent sigchld, but Python doesn't offer in Python < 3.4.  There is
+        # still a race condition here; we can get a sigchld while we're sitting in the waitpid call. However, AFAICT, if
+        # waitpid is interrupted by SIGCHLD, as long as we call waitpid again (which happens every so often during the
+        # normal course in the mainloop), we'll eventually reap the child that we tried to reap during the interrupted
+        # call. At least on Linux, this appears to be true, or at least stopping 50 processes at once never left zombies
+        # lying around.
+        try:
+            pid, sts = os.waitpid(-1, os.WNOHANG)
+        except OSError as exc:
+            code = exc.args[0]
+            if code not in (errno.ECHILD, errno.EINTR):
+                log.critical('waitpid error %r; a process may not be cleaned up properly', code)
+            if code == errno.EINTR:
+                log.debug('EINTR during reap')
+            pid, sts = None, None
+        return pid, sts
+
+    def set_uid_or_exit(self) -> None:
+        """
+        Set the uid of the supervisord process.  Called during supervisord startup only.  No return value.  Exits the
+        process via usage() if privileges could not be dropped.
+        """
+        if self.uid is None:
+            if os.getuid() == 0:
+                warnings.warn(
+                    'Supervisor is running as root.  Privileges were not dropped because no user is specified in the '
+                    'config file.  If you intend to run as root, you can set user=root in the config file to avoid '
+                    'this message.',
+                )
+        else:
+            msg = drop_privileges(self.uid)
+            if msg is None:
+                log.info('Set uid to user %s succeeded', self.uid)
+            else:  # failed to drop privileges
+                raise RuntimeError(msg)
+
+    def set_rlimits_or_exit(self) -> None:
+        """
+        Set the rlimits of the supervisord process.  Called during supervisord startup only.  No return value.  Exits
+        the process via usage() if any rlimits could not be set.
+        """
+
+        limits = []
+
+        if hasattr(resource, 'RLIMIT_NOFILE'):
+            limits.append({
+                'msg': (
+                    'The minimum number of file descriptors required to run this process is %(min_limit)s as per the '
+                    '"minfds" command-line argument or config file setting. The current environment will only allow '
+                    'you to open %(hard)s file descriptors.  Either raise the number of usable file descriptors in '
+                    'your environment (see README.rst) or lower the minfds setting in the config file to allow the '
+                    'process to start.'
+                ),
+                'min': self.config.minfds,
+                'resource': resource.RLIMIT_NOFILE,
+                'name': 'RLIMIT_NOFILE',
+            })
+
+        if hasattr(resource, 'RLIMIT_NPROC'):
+            limits.append({
+                'msg': (
+                    'The minimum number of available processes required to run this program is %(min_limit)s as per '
+                    'the "minprocs" command-line argument or config file setting. The current environment will only '
+                    'allow you to open %(hard)s processes.  Either raise the number of usable processes in your '
+                    'environment (see README.rst) or lower the minprocs setting in the config file to allow the '
+                    'program to start.'
+                ),
+                'min': self.config.minprocs,
+                'resource': resource.RLIMIT_NPROC,
+                'name': 'RLIMIT_NPROC',
+            })
+
+        for limit in limits:
+            min_limit = limit['min']
+            res = limit['resource']
+            msg = limit['msg']
+            name = limit['name']
+
+            soft, hard = resource.getrlimit(res)  # type: ignore
+
+            # -1 means unlimited
+            if soft < min_limit and soft != -1:  # type: ignore
+                if hard < min_limit and hard != -1:  # type: ignore
+                    # setrlimit should increase the hard limit if we are root, if not then setrlimit raises and we print
+                    # usage
+                    hard = min_limit  # type: ignore
+
+                try:
+                    resource.setrlimit(res, (min_limit, hard))  # type: ignore
+                    log.info('Increased %s limit to %s', name, min_limit)
+                except (resource.error, ValueError):
+                    raise RuntimeError(msg % dict(  # type: ignore  # noqa
+                        min_limit=min_limit,
+                        res=res,
+                        name=name,
+                        soft=soft,
+                        hard=hard,
+                    ))
+
+    def cleanup(self) -> None:
+        if self.unlink_pidfile:
+            try_unlink(self.config.pidfile)
+        self.poller.close()
+
+    def cleanup_fds(self) -> None:
+        # try to close any leaked file descriptors (for reload)
+        start = 5
+        os.closerange(start, self.config.minfds)
+
+    def clear_auto_child_logdir(self) -> None:
+        # must be called after realize()
+        child_logdir = self.config.child_logdir
+        fnre = re.compile(rf'.+?---{self.config.identifier}-\S+\.log\.?\d{{0,4}}')
+        try:
+            filenames = os.listdir(child_logdir)
+        except OSError:
+            log.warning('Could not clear child_log dir')
+            return
+
+        for filename in filenames:
+            if fnre.match(filename):
+                pathname = os.path.join(child_logdir, filename)
+                try:
+                    os.remove(pathname)
+                except OSError:
+                    log.warning('Failed to clean up %r', pathname)
+
+    def daemonize(self) -> None:
+        self.poller.before_daemonize()
+        self._daemonize()
+        self.poller.after_daemonize()
+
+    def _daemonize(self) -> None:
+        # To daemonize, we need to become the leader of our own session (process) group.  If we do not, signals sent to
+        # our parent process will also be sent to us.   This might be bad because signals such as SIGINT can be sent to
+        # our parent process during normal (uninteresting) operations such as when we press Ctrl-C in the parent
+        # terminal window to escape from a logtail command. To disassociate ourselves from our parent's session group we
+        # use os.setsid.  It means "set session id", which has the effect of disassociating a process from is current
+        # session and process group and setting itself up as a new session leader.
+        #
+        # Unfortunately we cannot call setsid if we're already a session group leader, so we use "fork" to make a copy
+        # of ourselves that is guaranteed to not be a session group leader.
+        #
+        # We also change directories, set stderr and stdout to null, and change our umask.
+        #
+        # This explanation was (gratefully) garnered from
+        # http://www.cems.uwe.ac.uk/~irjohnso/coursenotes/lrc/system/daemons/d3.htm
+
+        pid = os.fork()
+        if pid != 0:
+            # Parent
+            log.debug('supervisord forked; parent exiting')
+            real_exit(0)
+        # Child
+        log.info('daemonizing the supervisord process')
+        if self.config.directory:
+            try:
+                os.chdir(self.config.directory)
+            except OSError as err:
+                log.critical("can't chdir into %r: %s", self.config.directory, err)
+            else:
+                log.info('set current directory: %r', self.config.directory)
+        os.dup2(0, os.open('/dev/null', os.O_RDONLY))
+        os.dup2(1, os.open('/dev/null', os.O_WRONLY))
+        os.dup2(2, os.open('/dev/null', os.O_WRONLY))
+        os.setsid()
+        os.umask(self.config.umask)
+        # XXX Stevens, in his Advanced Unix book, section 13.3 (page 417) recommends calling umask(0) and closing unused
+        # file descriptors.  In his Network Programming book, he additionally recommends ignoring SIGHUP and forking
+        # again after the setsid() call, for obscure SVR4 reasons.
+
+    def get_auto_child_log_name(self, name: str, identifier: str, channel: str) -> str:
+        prefix = f'{name}-{channel}---{identifier}-'
+        logfile = mktempfile(
+            suffix='.log',
+            prefix=prefix,
+            dir=self.config.child_logdir,
+        )
+        return logfile
+
+    def get_signal(self) -> ta.Optional[int]:
+        return self.signal_receiver.get_signal()
+
+    def write_pidfile(self) -> None:
+        pid = os.getpid()
+        try:
+            with open(self.config.pidfile, 'w') as f:
+                f.write(f'{pid}\n')
+        except OSError:
+            log.critical('could not write pidfile %s', self.config.pidfile)
+        else:
+            self.unlink_pidfile = True
+            log.info('supervisord started with pid %s', pid)
+
+
+def drop_privileges(user: ta.Union[int, str, None]) -> ta.Optional[str]:
+    """
+    Drop privileges to become the specified user, which may be a username or uid.  Called for supervisord startup
+    and when spawning subprocesses.  Returns None on success or a string error message if privileges could not be
+    dropped.
+    """
+    if user is None:
+        return 'No user specified to setuid to!'
+
+    # get uid for user, which can be a number or username
+    try:
+        uid = int(user)
+    except ValueError:
+        try:
+            pwrec = pwd.getpwnam(user)  # type: ignore
+        except KeyError:
+            return f"Can't find username {user!r}"
+        uid = pwrec[2]
+    else:
+        try:
+            pwrec = pwd.getpwuid(uid)
+        except KeyError:
+            return f"Can't find uid {uid!r}"
+
+    current_uid = os.getuid()
+
+    if current_uid == uid:
+        # do nothing and return successfully if the uid is already the current one.  this allows a supervisord
+        # running as an unprivileged user "foo" to start a process where the config has "user=foo" (same user) in
+        # it.
+        return None
+
+    if current_uid != 0:
+        return "Can't drop privilege as nonroot user"
+
+    gid = pwrec[3]
+    if hasattr(os, 'setgroups'):
+        user = pwrec[0]
+        groups = [grprec[2] for grprec in grp.getgrall() if user in grprec[3]]
+
+        # always put our primary gid first in this list, otherwise we can lose group info since sometimes the first
+        # group in the setgroups list gets overwritten on the subsequent setgid call (at least on freebsd 9 with
+        # python 2.7 - this will be safe though for all unix /python version combos)
+        groups.insert(0, gid)
+        try:
+            os.setgroups(groups)
+        except OSError:
+            return 'Could not set groups of effective user'
+    try:
+        os.setgid(gid)
+    except OSError:
+        return 'Could not set group id of effective user'
+    os.setuid(uid)
+    return None
+
+
+def make_pipes(stderr=True) -> ta.Mapping[str, int]:
+    """
+    Create pipes for parent to child stdin/stdout/stderr communications.  Open fd in non-blocking mode so we can
+    read them in the mainloop without blocking.  If stderr is False, don't create a pipe for stderr.
+    """
+
+    pipes: ta.Dict[str, ta.Optional[int]] = {
+        'child_stdin': None,
+        'stdin': None,
+        'stdout': None,
+        'child_stdout': None,
+        'stderr': None,
+        'child_stderr': None,
+    }
+    try:
+        stdin, child_stdin = os.pipe()
+        pipes['child_stdin'], pipes['stdin'] = stdin, child_stdin
+        stdout, child_stdout = os.pipe()
+        pipes['stdout'], pipes['child_stdout'] = stdout, child_stdout
+        if stderr:
+            stderr, child_stderr = os.pipe()
+            pipes['stderr'], pipes['child_stderr'] = stderr, child_stderr
+        for fd in (pipes['stdout'], pipes['stderr'], pipes['stdin']):
+            if fd is not None:
+                flags = fcntl.fcntl(fd, fcntl.F_GETFL) | os.O_NDELAY
+                fcntl.fcntl(fd, fcntl.F_SETFL, flags)
+        return pipes  # type: ignore
+    except OSError:
+        for fd in pipes.values():
+            if fd is not None:
+                close_fd(fd)
+        raise
+
+
+def close_parent_pipes(pipes: ta.Mapping[str, int]) -> None:
+    for fdname in ('stdin', 'stdout', 'stderr'):
+        fd = pipes.get(fdname)
+        if fd is not None:
+            close_fd(fd)
+
+
+def close_child_pipes(pipes: ta.Mapping[str, int]) -> None:
+    for fdname in ('child_stdin', 'child_stdout', 'child_stderr'):
+        fd = pipes.get(fdname)
+        if fd is not None:
+            close_fd(fd)
+
+
+def check_execv_args(filename, argv, st) -> None:
+    if st is None:
+        raise NotFoundError(f"can't find command {filename!r}")
+
+    elif stat.S_ISDIR(st[stat.ST_MODE]):
+        raise NotExecutableError(f'command at {filename!r} is a directory')
+
+    elif not (stat.S_IMODE(st[stat.ST_MODE]) & 0o111):
+        raise NotExecutableError(f'command at {filename!r} is not executable')
+
+    elif not os.access(filename, os.X_OK):
+        raise NoPermissionError(f'no permission to run command {filename!r}')
+
+
+########################################
+# ../dispatchers.py
+
+
+log = logging.getLogger(__name__)
+
+
+class Dispatcher(abc.ABC):
+
+    def __init__(self, process: AbstractSubprocess, channel: str, fd: int) -> None:
+        super().__init__()
+
+        self._process = process  # process which "owns" this dispatcher
+        self._channel = channel  # 'stderr' or 'stdout'
+        self._fd = fd
+        self._closed = False  # True if close() has been called
+
+    def __repr__(self) -> str:
+        return f'<{self.__class__.__name__} at {id(self)} for {self._process} ({self._channel})>'
+
+    @property
+    def process(self) -> AbstractSubprocess:
+        return self._process
+
+    @property
+    def channel(self) -> str:
+        return self._channel
+
+    @property
+    def fd(self) -> int:
+        return self._fd
+
+    @property
+    def closed(self) -> bool:
+        return self._closed
+
+    @abc.abstractmethod
+    def readable(self) -> bool:
+        raise NotImplementedError
+
+    @abc.abstractmethod
+    def writable(self) -> bool:
+        raise NotImplementedError
+
+    def handle_read_event(self) -> None:
+        raise TypeError
+
+    def handle_write_event(self) -> None:
+        raise TypeError
+
+    def handle_error(self) -> None:
+        nil, t, v, tbinfo = compact_traceback()
+
+        log.critical('uncaptured python exception, closing channel %s (%s:%s %s)', repr(self), t, v, tbinfo)
+        self.close()
+
+    def close(self) -> None:
+        if not self._closed:
+            log.debug('fd %s closed, stopped monitoring %s', self._fd, self)
+            self._closed = True
+
+    def flush(self) -> None:  # noqa
+        pass
+
+
+class OutputDispatcher(Dispatcher):
+    """
+    Dispatcher for one channel (stdout or stderr) of one process. Serves several purposes:
+
+    - capture output sent within <!--XSUPERVISOR:BEGIN--> and <!--XSUPERVISOR:END--> tags and signal a
+      ProcessCommunicationEvent by calling notify_event(event).
+    - route the output to the appropriate log handlers as specified in the config.
+    """
+
+    child_log = None  # the current logger (normal_log or capture_log)
+    normal_log = None  # the "normal" (non-capture) logger
+    capture_log = None  # the logger used while we're in capture_mode
+    capture_mode = False  # are we capturing process event data
+    output_buffer = b''  # data waiting to be logged
+
+    def __init__(self, process: AbstractSubprocess, event_type, fd):
+        """
+        Initialize the dispatcher.
+
+        `event_type` should be one of ProcessLogStdoutEvent or ProcessLogStderrEvent
+        """
+        super().__init__(process, event_type.channel, fd)
+        self.event_type = event_type
+
+        self.lc: ProcessConfig.Log = getattr(process.config, self._channel)
+
+        self._init_normal_log()
+        self._init_capture_log()
+
+        self.child_log = self.normal_log
+
+        # all code below is purely for minor speedups
+        begin_token = self.event_type.BEGIN_TOKEN
+        end_token = self.event_type.END_TOKEN
+        self.begin_token_data = (begin_token, len(begin_token))
+        self.end_token_data = (end_token, len(end_token))
+        self.main_log_level = logging.DEBUG
+        config = self._process.config
+        self.log_to_main_log = process.context.config.loglevel <= self.main_log_level
+        self.stdout_events_enabled = config.stdout.events_enabled
+        self.stderr_events_enabled = config.stderr.events_enabled
+
+    def _init_normal_log(self) -> None:
+        """
+        Configure the "normal" (non-capture) log for this channel of this process. Sets self.normal_log if logging is
+        enabled.
+        """
+        config = self._process.config  # noqa
+        channel = self._channel  # noqa
+
+        logfile = self.lc.file
+        maxbytes = self.lc.maxbytes  # noqa
+        backups = self.lc.backups  # noqa
+        to_syslog = self.lc.syslog
+
+        if logfile or to_syslog:
+            self.normal_log = logging.getLogger(__name__)
+
+        # if logfile:
+        #     loggers.handle_file(
+        #         self.normal_log,
+        #         filename=logfile,
+        #         fmt='%(message)s',
+        #         rotating=bool(maxbytes),  # optimization
+        #         maxbytes=maxbytes,
+        #         backups=backups,
+        #     )
+        #
+        # if to_syslog:
+        #     loggers.handle_syslog(
+        #         self.normal_log,
+        #         fmt=config.name + ' %(message)s',
+        #     )
+
+    def _init_capture_log(self):
+        """
+        Configure the capture log for this process.  This log is used to temporarily capture output when special output
+        is detected. Sets self.capture_log if capturing is enabled.
+        """
+        capture_maxbytes = self.lc.capture_maxbytes
+        if capture_maxbytes:
+            self.capture_log = logging.getLogger(__name__)
+            # loggers.handle_boundIO(
+            #     self.capture_log,
+            #     fmt='%(message)s',
+            #     maxbytes=capture_maxbytes,
+            # )
+
+    def remove_logs(self):
+        for log in (self.normal_log, self.capture_log):
+            if log is not None:
+                for handler in log.handlers:
+                    handler.remove()  # type: ignore
+                    handler.reopen()  # type: ignore
+
+    def reopen_logs(self):
+        for log in (self.normal_log, self.capture_log):
+            if log is not None:
+                for handler in log.handlers:
+                    handler.reopen()  # type: ignore
+
+    def _log(self, data):
+        if data:
+            if self._process.context.config.strip_ansi:
+                data = strip_escapes(data)
+            if self.child_log:
+                self.child_log.info(data)  # type: ignore
+            if self.log_to_main_log:
+                if not isinstance(data, bytes):
+                    text = data
+                else:
+                    try:
+                        text = data.decode('utf-8')
+                    except UnicodeDecodeError:
+                        text = f'Undecodable: {data!r}'
+                log.log(self.main_log_level, '%r %s output:\n%s', self._process.config.name, self._channel, text)  # noqa
+            if self._channel == 'stdout':
+                if self.stdout_events_enabled:
+                    notify_event(ProcessLogStdoutEvent(self._process, self._process.pid, data))
+            elif self.stderr_events_enabled:
+                notify_event(ProcessLogStderrEvent(self._process, self._process.pid, data))
+
+    def record_output(self):
+        if self.capture_log is None:
+            # shortcut trying to find capture data
+            data = self.output_buffer
+            self.output_buffer = b''
+            self._log(data)
+            return
+
+        if self.capture_mode:
+            token, tokenlen = self.end_token_data
+        else:
+            token, tokenlen = self.begin_token_data
+
+        if len(self.output_buffer) <= tokenlen:
+            return  # not enough data
+
+        data = self.output_buffer
+        self.output_buffer = b''
+
+        try:
+            before, after = data.split(token, 1)
+        except ValueError:
+            after = None
+            index = find_prefix_at_end(data, token)
+            if index:
+                self.output_buffer = self.output_buffer + data[-index:]
+                data = data[:-index]
+            self._log(data)
+        else:
+            self._log(before)
+            self.toggle_capture_mode()
+            self.output_buffer = after  # type: ignore
+
+        if after:
+            self.record_output()
+
+    def toggle_capture_mode(self):
+        self.capture_mode = not self.capture_mode
+
+        if self.capture_log is not None:
+            if self.capture_mode:
+                self.child_log = self.capture_log
+            else:
+                for handler in self.capture_log.handlers:
+                    handler.flush()
+                data = self.capture_log.getvalue()  # type: ignore
+                channel = self._channel
+                procname = self._process.config.name
+                event = self.event_type(self._process, self._process.pid, data)
+                notify_event(event)
+
+                log.debug('%r %s emitted a comm event', procname, channel)
+                for handler in self.capture_log.handlers:
+                    handler.remove()  # type: ignore
+                    handler.reopen()  # type: ignore
+                self.child_log = self.normal_log
+
+    def writable(self) -> bool:
+        return False
+
+    def readable(self) -> bool:
+        if self._closed:
+            return False
+        return True
+
+    def handle_read_event(self) -> None:
+        data = readfd(self._fd)
+        self.output_buffer += data
+        self.record_output()
+        if not data:
+            # if we get no data back from the pipe, it means that the child process has ended.  See
+            # mail.python.org/pipermail/python-dev/2004-August/046850.html
+            self.close()
+
+
+class InputDispatcher(Dispatcher):
+
+    def __init__(self, process: AbstractSubprocess, channel: str, fd: int) -> None:
+        super().__init__(process, channel, fd)
+        self._input_buffer = b''
+
+    def writable(self) -> bool:
+        if self._input_buffer and not self._closed:
+            return True
+        return False
+
+    def readable(self) -> bool:
+        return False
+
+    def flush(self) -> None:
+        # other code depends on this raising EPIPE if the pipe is closed
+        sent = os.write(self._fd, as_bytes(self._input_buffer))
+        self._input_buffer = self._input_buffer[sent:]
+
+    def handle_write_event(self) -> None:
+        if self._input_buffer:
+            try:
+                self.flush()
+            except OSError as why:
+                if why.args[0] == errno.EPIPE:
+                    self._input_buffer = b''
+                    self.close()
+                else:
+                    raise
+
+
+########################################
+# ../process.py
+
+
+log = logging.getLogger(__name__)
+
+
+@functools.total_ordering
+class Subprocess(AbstractSubprocess):
+    """A class to manage a subprocess."""
+
+    # Initial state; overridden by instance variables
+
+    # pid = 0  # Subprocess pid; 0 when not running
+    # config = None  # ProcessConfig instance
+    # state = None  # process state code
+    listener_state = None  # listener state code (if we're an event listener)
+    event = None  # event currently being processed (if we're an event listener)
+    laststart = 0.  # Last time the subprocess was started; 0 if never
+    laststop = 0.  # Last time the subprocess was stopped; 0 if never
+    last_stop_report = 0.  # Last time "waiting for x to stop" logged, to throttle
+    delay = 0.  # If nonzero, delay starting or killing until this time
+    administrative_stop = False  # true if process has been stopped by an admin
+    system_stop = False  # true if process has been stopped by the system
+    killing = False  # true if we are trying to kill this process
+    backoff = 0  # backoff counter (to startretries)
+    dispatchers = None  # asyncore output dispatchers (keyed by fd)
+    pipes = None  # map of channel name to file descriptor #
+    exitstatus = None  # status attached to dead process by finish()
+    spawn_err = None  # error message attached by spawn() if any
+    group = None  # ProcessGroup instance if process is in the group
+
+    def __init__(self, config: ProcessConfig, group: 'ProcessGroup', context: AbstractServerContext) -> None:
+        super().__init__()
+        self._config = config
+        self.group = group
+        self._context = context
+        self._dispatchers: dict = {}
+        self._pipes: dict = {}
+        self.state = ProcessStates.STOPPED
+        self._pid = 0
+
+    @property
+    def pid(self) -> int:
+        return self._pid
+
+    @property
+    def config(self) -> ProcessConfig:
+        return self._config
+
+    @property
+    def context(self) -> AbstractServerContext:
+        return self._context
+
+    def remove_logs(self) -> None:
+        for dispatcher in self._dispatchers.values():
+            if hasattr(dispatcher, 'remove_logs'):
+                dispatcher.remove_logs()
+
+    def reopen_logs(self) -> None:
+        for dispatcher in self._dispatchers.values():
+            if hasattr(dispatcher, 'reopen_logs'):
+                dispatcher.reopen_logs()
+
+    def drain(self) -> None:
+        for dispatcher in self._dispatchers.values():
+            # note that we *must* call readable() for every dispatcher, as it may have side effects for a given
+            # dispatcher (eg. call handle_listener_state_change for event listener processes)
+            if dispatcher.readable():
+                dispatcher.handle_read_event()
+            if dispatcher.writable():
+                dispatcher.handle_write_event()
+
+    def write(self, chars: ta.Union[bytes, str]) -> None:
+        if not self.pid or self.killing:
+            raise OSError(errno.EPIPE, 'Process already closed')
+
+        stdin_fd = self._pipes['stdin']
+        if stdin_fd is None:
+            raise OSError(errno.EPIPE, 'Process has no stdin channel')
+
+        dispatcher = self._dispatchers[stdin_fd]
+        if dispatcher.closed:
+            raise OSError(errno.EPIPE, "Process' stdin channel is closed")
+
+        dispatcher.input_buffer += chars
+        dispatcher.flush()  # this must raise EPIPE if the pipe is closed
+
+    def _get_execv_args(self) -> ta.Tuple[str, ta.Sequence[str]]:
+        """
+        Internal: turn a program name into a file name, using $PATH, make sure it exists / is executable, raising a
+        ProcessError if not
+        """
+        try:
+            commandargs = shlex.split(self.config.command)
+        except ValueError as e:
+            raise BadCommandError(f"can't parse command {self.config.command!r}: {e}")  # noqa
+
+        if commandargs:
+            program = commandargs[0]
+        else:
+            raise BadCommandError('command is empty')
+
+        if '/' in program:
+            filename = program
+            try:
+                st = os.stat(filename)
+            except OSError:
+                st = None
+
+        else:
+            path = get_path()
+            found = None
+            st = None
+            for dir in path:  # noqa
+                found = os.path.join(dir, program)
+                try:
+                    st = os.stat(found)
+                except OSError:
+                    pass
+                else:
+                    break
+            if st is None:
+                filename = program
+            else:
+                filename = found  # type: ignore
+
+        # check_execv_args will raise a ProcessError if the execv args are bogus, we break it out into a separate
+        # options method call here only to service unit tests
+        check_execv_args(filename, commandargs, st)
+
+        return filename, commandargs
+
+    event_map: ta.ClassVar[ta.Mapping[int, ta.Type[ProcessStateEvent]]] = {
+        ProcessStates.BACKOFF: ProcessStateBackoffEvent,
+        ProcessStates.FATAL: ProcessStateFatalEvent,
+        ProcessStates.UNKNOWN: ProcessStateUnknownEvent,
+        ProcessStates.STOPPED: ProcessStateStoppedEvent,
+        ProcessStates.EXITED: ProcessStateExitedEvent,
+        ProcessStates.RUNNING: ProcessStateRunningEvent,
+        ProcessStates.STARTING: ProcessStateStartingEvent,
+        ProcessStates.STOPPING: ProcessStateStoppingEvent,
+    }
+
+    def change_state(self, new_state: ProcessState, expected: bool = True) -> bool:
+        old_state = self.state
+        if new_state is old_state:
+            return False
+
+        self.state = new_state
+        if new_state == ProcessStates.BACKOFF:
+            now = time.time()
+            self.backoff += 1
+            self.delay = now + self.backoff
+
+        event_class = self.event_map.get(new_state)
+        if event_class is not None:
+            event = event_class(self, old_state, expected)
+            notify_event(event)
+
+        return True
+
+    def _check_in_state(self, *states: ProcessState) -> None:
+        if self.state not in states:
+            current_state = get_process_state_description(self.state)
+            allowable_states = ' '.join(map(get_process_state_description, states))
+            processname = as_string(self.config.name)
+            raise AssertionError('Assertion failed for %s: %s not in %s' % (processname, current_state, allowable_states))  # noqa
+
+    def _record_spawn_err(self, msg: str) -> None:
+        self.spawn_err = msg
+        log.info('spawn_err: %s', msg)
+
+    def spawn(self) -> ta.Optional[int]:
+        processname = as_string(self.config.name)
+
+        if self.pid:
+            log.warning('process \'%s\' already running', processname)
+            return None
+
+        self.killing = False
+        self.spawn_err = None
+        self.exitstatus = None
+        self.system_stop = False
+        self.administrative_stop = False
+
+        self.laststart = time.time()
+
+        self._check_in_state(
+            ProcessStates.EXITED,
+            ProcessStates.FATAL,
+            ProcessStates.BACKOFF,
+            ProcessStates.STOPPED,
+        )
+
+        self.change_state(ProcessStates.STARTING)
+
+        try:
+            filename, argv = self._get_execv_args()
+        except ProcessError as what:
+            self._record_spawn_err(what.args[0])
+            self._check_in_state(ProcessStates.STARTING)
+            self.change_state(ProcessStates.BACKOFF)
+            return None
+
+        try:
+            self._dispatchers, self._pipes = self._make_dispatchers()  # type: ignore
+        except OSError as why:
+            code = why.args[0]
+            if code == errno.EMFILE:
+                # too many file descriptors open
+                msg = f"too many open files to spawn '{processname}'"
+            else:
+                msg = f"unknown error making dispatchers for '{processname}': {errno.errorcode.get(code, code)}"
+            self._record_spawn_err(msg)
+            self._check_in_state(ProcessStates.STARTING)
+            self.change_state(ProcessStates.BACKOFF)
+            return None
+
+        try:
+            pid = os.fork()
+        except OSError as why:
+            code = why.args[0]
+            if code == errno.EAGAIN:
+                # process table full
+                msg = f'Too many processes in process table to spawn \'{processname}\''
+            else:
+                msg = f'unknown error during fork for \'{processname}\': {errno.errorcode.get(code, code)}'
+            self._record_spawn_err(msg)
+            self._check_in_state(ProcessStates.STARTING)
+            self.change_state(ProcessStates.BACKOFF)
+            close_parent_pipes(self._pipes)
+            close_child_pipes(self._pipes)
+            return None
+
+        if pid != 0:
+            return self._spawn_as_parent(pid)
+
+        else:
+            self._spawn_as_child(filename, argv)
+            return None
+
+    def _make_dispatchers(self) -> ta.Tuple[ta.Mapping[int, Dispatcher], ta.Mapping[str, int]]:
+        use_stderr = not self.config.redirect_stderr
+        p = make_pipes(use_stderr)
+        stdout_fd, stderr_fd, stdin_fd = p['stdout'], p['stderr'], p['stdin']
+        dispatchers: ta.Dict[int, Dispatcher] = {}
+        etype: ta.Type[ProcessCommunicationEvent]
+        if stdout_fd is not None:
+            etype = ProcessCommunicationStdoutEvent
+            dispatchers[stdout_fd] = OutputDispatcher(self, etype, stdout_fd)
+        if stderr_fd is not None:
+            etype = ProcessCommunicationStderrEvent
+            dispatchers[stderr_fd] = OutputDispatcher(self, etype, stderr_fd)
+        if stdin_fd is not None:
+            dispatchers[stdin_fd] = InputDispatcher(self, 'stdin', stdin_fd)
+        return dispatchers, p
+
+    def _spawn_as_parent(self, pid: int) -> int:
+        # Parent
+        self._pid = pid
+        close_child_pipes(self._pipes)
+        log.info('spawned: \'%s\' with pid %s', as_string(self.config.name), pid)
+        self.spawn_err = None
+        self.delay = time.time() + self.config.startsecs
+        self.context.pid_history[pid] = self
+        return pid
+
+    def _prepare_child_fds(self) -> None:
+        os.dup2(self._pipes['child_stdin'], 0)
+        os.dup2(self._pipes['child_stdout'], 1)
+        if self.config.redirect_stderr:
+            os.dup2(self._pipes['child_stdout'], 2)
+        else:
+            os.dup2(self._pipes['child_stderr'], 2)
+        for i in range(3, self.context.config.minfds):
+            close_fd(i)
+
+    def _spawn_as_child(self, filename: str, argv: ta.Sequence[str]) -> None:
+        try:
+            # prevent child from receiving signals sent to the parent by calling os.setpgrp to create a new process
+            # group for the child; this prevents, for instance, the case of child processes being sent a SIGINT when
+            # running supervisor in foreground mode and Ctrl-C in the terminal window running supervisord is pressed.
+            # Presumably it also prevents HUP, etc received by supervisord from being sent to children.
+            os.setpgrp()
+
+            self._prepare_child_fds()
+            # sending to fd 2 will put this output in the stderr log
+
+            # set user
+            setuid_msg = self.set_uid()
+            if setuid_msg:
+                uid = self.config.uid
+                msg = f"couldn't setuid to {uid}: {setuid_msg}\n"
+                os.write(2, as_bytes('supervisor: ' + msg))
+                return  # finally clause will exit the child process
+
+            # set environment
+            env = os.environ.copy()
+            env['SUPERVISOR_ENABLED'] = '1'
+            env['SUPERVISOR_PROCESS_NAME'] = self.config.name
+            if self.group:
+                env['SUPERVISOR_GROUP_NAME'] = self.group.config.name
+            if self.config.environment is not None:
+                env.update(self.config.environment)
+
+            # change directory
+            cwd = self.config.directory
+            try:
+                if cwd is not None:
+                    os.chdir(cwd)
+            except OSError as why:
+                code = errno.errorcode.get(why.args[0], why.args[0])
+                msg = f"couldn't chdir to {cwd}: {code}\n"
+                os.write(2, as_bytes('supervisor: ' + msg))
+                return  # finally clause will exit the child process
+
+            # set umask, then execve
+            try:
+                if self.config.umask is not None:
+                    os.umask(self.config.umask)
+                os.execve(filename, list(argv), env)
+            except OSError as why:
+                code = errno.errorcode.get(why.args[0], why.args[0])
+                msg = f"couldn't exec {argv[0]}: {code}\n"
+                os.write(2, as_bytes('supervisor: ' + msg))
+            except Exception:  # noqa
+                (file, fun, line), t, v, tbinfo = compact_traceback()
+                error = f'{t}, {v}: file: {file} line: {line}'
+                msg = f"couldn't exec {filename}: {error}\n"
+                os.write(2, as_bytes('supervisor: ' + msg))
+
+            # this point should only be reached if execve failed. the finally clause will exit the child process.
+
+        finally:
+            os.write(2, as_bytes('supervisor: child process was not spawned\n'))
+            real_exit(127)  # exit process with code for spawn failure
+
+    def _check_and_adjust_for_system_clock_rollback(self, test_time):
+        """
+        Check if system clock has rolled backward beyond test_time. If so, set affected timestamps to test_time.
+        """
+        if self.state == ProcessStates.STARTING:
+            self.laststart = min(test_time, self.laststart)
+            if self.delay > 0 and test_time < (self.delay - self.config.startsecs):
+                self.delay = test_time + self.config.startsecs
+
+        elif self.state == ProcessStates.RUNNING:
+            if test_time > self.laststart and test_time < (self.laststart + self.config.startsecs):
+                self.laststart = test_time - self.config.startsecs
+
+        elif self.state == ProcessStates.STOPPING:
+            self.last_stop_report = min(test_time, self.last_stop_report)
+            if self.delay > 0 and test_time < (self.delay - self.config.stopwaitsecs):
+                self.delay = test_time + self.config.stopwaitsecs
+
+        elif self.state == ProcessStates.BACKOFF:
+            if self.delay > 0 and test_time < (self.delay - self.backoff):
+                self.delay = test_time + self.backoff
+
+    def stop(self) -> ta.Optional[str]:
+        self.administrative_stop = True
+        self.last_stop_report = 0
+        return self.kill(self.config.stopsignal)
+
+    def stop_report(self) -> None:
+        """ Log a 'waiting for x to stop' message with throttling. """
+        if self.state == ProcessStates.STOPPING:
+            now = time.time()
+
+            self._check_and_adjust_for_system_clock_rollback(now)
+
+            if now > (self.last_stop_report + 2):  # every 2 seconds
+                log.info('waiting for %s to stop', as_string(self.config.name))
+                self.last_stop_report = now
+
+    def give_up(self) -> None:
+        self.delay = 0
+        self.backoff = 0
+        self.system_stop = True
+        self._check_in_state(ProcessStates.BACKOFF)
+        self.change_state(ProcessStates.FATAL)
+
+    def kill(self, sig: int) -> ta.Optional[str]:
+        """
+        Send a signal to the subprocess with the intention to kill it (to make it exit).  This may or may not actually
+        kill it.
+
+        Return None if the signal was sent, or an error message string if an error occurred or if the subprocess is not
+        running.
+        """
+        now = time.time()
+
+        processname = as_string(self.config.name)
+        # If the process is in BACKOFF and we want to stop or kill it, then BACKOFF -> STOPPED.  This is needed because
+        # if startretries is a large number and the process isn't starting successfully, the stop request would be
+        # blocked for a long time waiting for the retries.
+        if self.state == ProcessStates.BACKOFF:
+            log.debug('Attempted to kill %s, which is in BACKOFF state.', processname)
+            self.change_state(ProcessStates.STOPPED)
+            return None
+
+        args: tuple
+        if not self.pid:
+            fmt, args = "attempted to kill %s with sig %s but it wasn't running", (processname, signame(sig))
+            log.debug(fmt, *args)
+            return fmt % args
+
+        # If we're in the stopping state, then we've already sent the stop signal and this is the kill signal
+        if self.state == ProcessStates.STOPPING:
+            killasgroup = self.config.killasgroup
+        else:
+            killasgroup = self.config.stopasgroup
+
+        as_group = ''
+        if killasgroup:
+            as_group = 'process group '
+
+        log.debug('killing %s (pid %s) %swith signal %s', processname, self.pid, as_group, signame(sig))
+
+        # RUNNING/STARTING/STOPPING -> STOPPING
+        self.killing = True
+        self.delay = now + self.config.stopwaitsecs
+        # we will already be in the STOPPING state if we're doing a SIGKILL as a result of overrunning stopwaitsecs
+        self._check_in_state(ProcessStates.RUNNING, ProcessStates.STARTING, ProcessStates.STOPPING)
+        self.change_state(ProcessStates.STOPPING)
+
+        pid = self.pid
+        if killasgroup:
+            # send to the whole process group instead
+            pid = -self.pid
+
+        try:
+            try:
+                os.kill(pid, sig)
+            except OSError as exc:
+                if exc.errno == errno.ESRCH:
+                    log.debug('unable to signal %s (pid %s), it probably just exited on its own: %s', processname, self.pid, str(exc))  # noqa
+                    # we could change the state here but we intentionally do not.  we will do it during normal SIGCHLD
+                    # processing.
+                    return None
+                raise
+        except Exception:  # noqa
+            tb = traceback.format_exc()
+            fmt, args = 'unknown problem killing %s (%s):%s', (processname, self.pid, tb)
+            log.critical(fmt, *args)
+            self.change_state(ProcessStates.UNKNOWN)
+            self.killing = False
+            self.delay = 0
+            return fmt % args
+
+        return None
+
+    def signal(self, sig: int) -> ta.Optional[str]:
+        """
+        Send a signal to the subprocess, without intending to kill it.
+
+        Return None if the signal was sent, or an error message string if an error occurred or if the subprocess is not
+        running.
+        """
+        processname = as_string(self.config.name)
+        args: tuple
+        if not self.pid:
+            fmt, args = "attempted to send %s sig %s but it wasn't running", (processname, signame(sig))
+            log.debug(fmt, *args)
+            return fmt % args
+
+        log.debug('sending %s (pid %s) sig %s', processname, self.pid, signame(sig))
+
+        self._check_in_state(ProcessStates.RUNNING, ProcessStates.STARTING, ProcessStates.STOPPING)
+
+        try:
+            try:
+                os.kill(self.pid, sig)
+            except OSError as exc:
+                if exc.errno == errno.ESRCH:
+                    log.debug('unable to signal %s (pid %s), it probably just now exited '
+                              'on its own: %s', processname, self.pid, str(exc))
+                    # we could change the state here but we intentionally do not.  we will do it during normal SIGCHLD
+                    # processing.
+                    return None
+                raise
+        except Exception:  # noqa
+            tb = traceback.format_exc()
+            fmt, args = 'unknown problem sending sig %s (%s):%s', (processname, self.pid, tb)
+            log.critical(fmt, *args)
+            self.change_state(ProcessStates.UNKNOWN)
+            return fmt % args
+
+        return None
+
+    def finish(self, sts: int) -> None:
+        """ The process was reaped and we need to report and manage its state """
+        self.drain()
+
+        es, msg = decode_wait_status(sts)
+
+        now = time.time()
+
+        self._check_and_adjust_for_system_clock_rollback(now)
+
+        self.laststop = now
+        processname = as_string(self.config.name)
+
+        if now > self.laststart:
+            too_quickly = now - self.laststart < self.config.startsecs
+        else:
+            too_quickly = False
+            log.warning(
+                "process '%s' (%s) laststart time is in the future, don't "
+                "know how long process was running so assuming it did "
+                "not exit too quickly", processname, self.pid)
+
+        exit_expected = es in self.config.exitcodes
+
+        if self.killing:
+            # likely the result of a stop request implies STOPPING -> STOPPED
+            self.killing = False
+            self.delay = 0
+            self.exitstatus = es
+
+            fmt, args = 'stopped: %s (%s)', (processname, msg)
+            self._check_in_state(ProcessStates.STOPPING)
+            self.change_state(ProcessStates.STOPPED)
+            if exit_expected:
+                log.info(fmt, *args)
+            else:
+                log.warning(fmt, *args)
+
+        elif too_quickly:
+            # the program did not stay up long enough to make it to RUNNING implies STARTING -> BACKOFF
+            self.exitstatus = None
+            self.spawn_err = 'Exited too quickly (process log may have details)'
+            self._check_in_state(ProcessStates.STARTING)
+            self.change_state(ProcessStates.BACKOFF)
+            log.warning('exited: %s (%s)', processname, msg + '; not expected')
+
+        else:
+            # this finish was not the result of a stop request, the program was in the RUNNING state but exited implies
+            # RUNNING -> EXITED normally but see next comment
+            self.delay = 0
+            self.backoff = 0
+            self.exitstatus = es
+
+            # if the process was STARTING but a system time change causes self.laststart to be in the future, the normal
+            # STARTING->RUNNING transition can be subverted so we perform the transition here.
+            if self.state == ProcessStates.STARTING:
+                self.change_state(ProcessStates.RUNNING)
+
+            self._check_in_state(ProcessStates.RUNNING)
+
+            if exit_expected:
+                # expected exit code
+                self.change_state(ProcessStates.EXITED, expected=True)
+                log.info('exited: %s (%s)', processname, msg + '; expected')
+            else:
+                # unexpected exit code
+                self.spawn_err = f'Bad exit code {es}'
+                self.change_state(ProcessStates.EXITED, expected=False)
+                log.warning('exited: %s (%s)', processname, msg + '; not expected')
+
+        self._pid = 0
+        close_parent_pipes(self._pipes)
+        self._pipes = {}
+        self._dispatchers = {}
+
+        # if we died before we processed the current event (only happens if we're an event listener), notify the event
+        # system that this event was rejected so it can be processed again.
+        if self.event is not None:
+            # Note: this should only be true if we were in the BUSY state when finish() was called.
+            notify_event(EventRejectedEvent(self, self.event))  # type: ignore
+            self.event = None
+
+    def set_uid(self) -> ta.Optional[str]:
+        if self.config.uid is None:
+            return None
+        msg = drop_privileges(self.config.uid)
+        return msg
+
+    def __lt__(self, other):
+        return self.config.priority < other.config.priority
+
+    def __eq__(self, other):
+        return self.config.priority == other.config.priority
+
+    def __repr__(self):
+        # repr can't return anything other than a native string, but the name might be unicode - a problem on Python 2.
+        name = self.config.name
+        return f'<Subprocess at {id(self)} with name {name} in state {get_process_state_description(self.get_state())}>'
+
+    def get_state(self) -> ProcessState:
+        return self.state
+
+    def transition(self):
+        now = time.time()
+        state = self.state
+
+        self._check_and_adjust_for_system_clock_rollback(now)
+
+        logger = log
+
+        if self.context.state > SupervisorStates.RESTARTING:
+            # dont start any processes if supervisor is shutting down
+            if state == ProcessStates.EXITED:
+                if self.config.autorestart:
+                    if self.config.autorestart is RestartUnconditionally:
+                        # EXITED -> STARTING
+                        self.spawn()
+                    elif self.exitstatus not in self.config.exitcodes:  # type: ignore
+                        # EXITED -> STARTING
+                        self.spawn()
+
+            elif state == ProcessStates.STOPPED and not self.laststart:
+                if self.config.autostart:
+                    # STOPPED -> STARTING
+                    self.spawn()
+
+            elif state == ProcessStates.BACKOFF:
+                if self.backoff <= self.config.startretries:
+                    if now > self.delay:
+                        # BACKOFF -> STARTING
+                        self.spawn()
+
+        processname = as_string(self.config.name)
+        if state == ProcessStates.STARTING:
+            if now - self.laststart > self.config.startsecs:
+                # STARTING -> RUNNING if the proc has started successfully and it has stayed up for at least
+                # proc.config.startsecs,
+                self.delay = 0
+                self.backoff = 0
+                self._check_in_state(ProcessStates.STARTING)
+                self.change_state(ProcessStates.RUNNING)
+                msg = ('entered RUNNING state, process has stayed up for > than %s seconds (startsecs)' % self.config.startsecs)  # noqa
+                logger.info('success: %s %s', processname, msg)
+
+        if state == ProcessStates.BACKOFF:
+            if self.backoff > self.config.startretries:
+                # BACKOFF -> FATAL if the proc has exceeded its number of retries
+                self.give_up()
+                msg = ('entered FATAL state, too many start retries too quickly')
+                logger.info('gave up: %s %s', processname, msg)
+
+        elif state == ProcessStates.STOPPING:
+            time_left = self.delay - now
+            if time_left <= 0:
+                # kill processes which are taking too long to stop with a final sigkill.  if this doesn't kill it, the
+                # process will be stuck in the STOPPING state forever.
+                log.warning('killing \'%s\' (%s) with SIGKILL', processname, self.pid)
+                self.kill(signal.SIGKILL)
+
+    def create_auto_child_logs(self):
+        # temporary logfiles which are erased at start time
+        # get_autoname = self.context.get_auto_child_log_name  # noqa
+        # sid = self.context.config.identifier  # noqa
+        # name = self.config.name  # noqa
+        # if self.stdout_logfile is Automatic:
+        #     self.stdout_logfile = get_autoname(name, sid, 'stdout')
+        # if self.stderr_logfile is Automatic:
+        #     self.stderr_logfile = get_autoname(name, sid, 'stderr')
+        pass
+
+
+@functools.total_ordering
+class ProcessGroup:
+    def __init__(self, config: ProcessGroupConfig, context: ServerContext):
+        super().__init__()
+        self.config = config
+        self.context = context
+        self.processes = {}
+        for pconfig in self.config.processes or []:
+            process = Subprocess(pconfig, self, self.context)
+            self.processes[pconfig.name] = process
+
+    def __lt__(self, other):
+        return self.config.priority < other.config.priority
+
+    def __eq__(self, other):
+        return self.config.priority == other.config.priority
+
+    def __repr__(self):
+        # repr can't return anything other than a native string, but the name might be unicode - a problem on Python 2.
+        name = self.config.name
+        return f'<{self.__class__.__name__} instance at {id(self)} named {name}>'
+
+    def remove_logs(self) -> None:
+        for process in self.processes.values():
+            process.remove_logs()
+
+    def reopen_logs(self) -> None:
+        for process in self.processes.values():
+            process.reopen_logs()
+
+    def stop_all(self) -> None:
+        processes = list(self.processes.values())
+        processes.sort()
+        processes.reverse()  # stop in desc priority order
+
+        for proc in processes:
+            state = proc.get_state()
+            if state == ProcessStates.RUNNING:
+                # RUNNING -> STOPPING
+                proc.stop()
+
+            elif state == ProcessStates.STARTING:
+                # STARTING -> STOPPING
+                proc.stop()
+
+            elif state == ProcessStates.BACKOFF:
+                # BACKOFF -> FATAL
+                proc.give_up()
+
+    def get_unstopped_processes(self) -> ta.List[Subprocess]:
+        return [x for x in self.processes.values() if x.get_state() not in STOPPED_STATES]
+
+    def get_dispatchers(self) -> ta.Dict[int, Dispatcher]:
+        dispatchers = {}
+        for process in self.processes.values():
+            dispatchers.update(process._dispatchers)  # noqa
+        return dispatchers
+
+    def before_remove(self) -> None:
+        pass
+
+    def transition(self) -> None:
+        for proc in self.processes.values():
+            proc.transition()
+
+    def after_setuid(self) -> None:
+        for proc in self.processes.values():
+            proc.create_auto_child_logs()
+
+
+########################################
+# supervisor.py
+
+
+log = logging.getLogger(__name__)
+
+
+class Supervisor:
+
+    def __init__(self, context: ServerContext) -> None:
+        super().__init__()
+
+        self._context = context
+        self._ticks: ta.Dict[int, float] = {}
+        self._process_groups: ta.Dict[str, ProcessGroup] = {}  # map of process group name to process group object
+        self._stop_groups: ta.Optional[ta.List[ProcessGroup]] = None  # list used for priority ordered shutdown
+        self._stopping = False  # set after we detect that we are handling a stop request
+        self._last_shutdown_report = 0.  # throttle for delayed process error reports at stop
+
+    @property
+    def context(self) -> ServerContext:
+        return self._context
+
+    def get_state(self) -> SupervisorState:
+        return self._context.state
+
+    def main(self) -> None:
+        if not self._context.first:
+            # prevent crash on libdispatch-based systems, at least for the first request
+            self._context.cleanup_fds()
+
+        self._context.set_uid_or_exit()
+
+        if self._context.first:
+            self._context.set_rlimits_or_exit()
+
+        # this sets the options.logger object delay logger instantiation until after setuid
+        if not self._context.config.nocleanup:
+            # clean up old automatic logs
+            self._context.clear_auto_child_logdir()
+
+        self.run()
+
+    def run(self) -> None:
+        self._process_groups = {}  # clear
+        self._stop_groups = None  # clear
+
+        clear_events()
+
+        try:
+            for config in self._context.config.groups or []:
+                self.add_process_group(config)
+
+            self._context.set_signals()
+
+            if not self._context.config.nodaemon and self._context.first:
+                self._context.daemonize()
+
+            # writing pid file needs to come *after* daemonizing or pid will be wrong
+            self._context.write_pidfile()
+
+            self.runforever()
+
+        finally:
+            self._context.cleanup()
+
+    def diff_to_active(self):
+        new = self._context.config.groups or []
+        cur = [group.config for group in self._process_groups.values()]
+
+        curdict = dict(zip([cfg.name for cfg in cur], cur))
+        newdict = dict(zip([cfg.name for cfg in new], new))
+
+        added = [cand for cand in new if cand.name not in curdict]
+        removed = [cand for cand in cur if cand.name not in newdict]
+
+        changed = [cand for cand in new if cand != curdict.get(cand.name, cand)]
+
+        return added, changed, removed
+
+    def add_process_group(self, config: ProcessGroupConfig) -> bool:
+        name = config.name
+        if name in self._process_groups:
+            return False
+
+        group = self._process_groups[name] = ProcessGroup(config, self._context)
+        group.after_setuid()
+
+        notify_event(ProcessGroupAddedEvent(name))
+        return True
+
+    def remove_process_group(self, name: str) -> bool:
+        if self._process_groups[name].get_unstopped_processes():
+            return False
+
+        self._process_groups[name].before_remove()
+
+        del self._process_groups[name]
+
+        notify_event(ProcessGroupRemovedEvent(name))
+        return True
+
+    def get_process_map(self) -> ta.Dict[int, Dispatcher]:
+        process_map = {}
+        for group in self._process_groups.values():
+            process_map.update(group.get_dispatchers())
+        return process_map
+
+    def shutdown_report(self) -> ta.List[Subprocess]:
+        unstopped: ta.List[Subprocess] = []
+
+        for group in self._process_groups.values():
+            unstopped.extend(group.get_unstopped_processes())
+
+        if unstopped:
+            # throttle 'waiting for x to die' reports
+            now = time.time()
+            if now > (self._last_shutdown_report + 3):  # every 3 secs
+                names = [as_string(p.config.name) for p in unstopped]
+                namestr = ', '.join(names)
+                log.info('waiting for %s to die', namestr)
+                self._last_shutdown_report = now
+                for proc in unstopped:
+                    state = get_process_state_description(proc.get_state())
+                    log.debug('%s state: %s', proc.config.name, state)
+
+        return unstopped
+
+    def _ordered_stop_groups_phase_1(self) -> None:
+        if self._stop_groups:
+            # stop the last group (the one with the "highest" priority)
+            self._stop_groups[-1].stop_all()
+
+    def _ordered_stop_groups_phase_2(self) -> None:
+        # after phase 1 we've transitioned and reaped, let's see if we can remove the group we stopped from the
+        # stop_groups queue.
+        if self._stop_groups:
+            # pop the last group (the one with the "highest" priority)
+            group = self._stop_groups.pop()
+            if group.get_unstopped_processes():
+                # if any processes in the group aren't yet in a stopped state, we're not yet done shutting this group
+                # down, so push it back on to the end of the stop group queue
+                self._stop_groups.append(group)
+
+    def runforever(self) -> None:
+        notify_event(SupervisorRunningEvent())
+        timeout = 1  # this cannot be fewer than the smallest TickEvent (5)
+
+        while True:
+            combined_map = {}
+            combined_map.update(self.get_process_map())
+
+            pgroups = list(self._process_groups.values())
+            pgroups.sort()
+
+            if self._context.state < SupervisorStates.RUNNING:
+                if not self._stopping:
+                    # first time, set the stopping flag, do a notification and set stop_groups
+                    self._stopping = True
+                    self._stop_groups = pgroups[:]
+                    notify_event(SupervisorStoppingEvent())
+
+                self._ordered_stop_groups_phase_1()
+
+                if not self.shutdown_report():
+                    # if there are no unstopped processes (we're done killing everything), it's OK to shutdown or reload
+                    raise ExitNow
+
+            for fd, dispatcher in combined_map.items():
+                if dispatcher.readable():
+                    self._context.poller.register_readable(fd)
+                if dispatcher.writable():
+                    self._context.poller.register_writable(fd)
+
+            r, w = self._context.poller.poll(timeout)
+
+            for fd in r:
+                if fd in combined_map:
+                    try:
+                        dispatcher = combined_map[fd]
+                        log.debug('read event caused by %r', dispatcher)
+                        dispatcher.handle_read_event()
+                        if not dispatcher.readable():
+                            self._context.poller.unregister_readable(fd)
+                    except ExitNow:
+                        raise
+                    except Exception:  # noqa
+                        combined_map[fd].handle_error()
+                else:
+                    # if the fd is not in combined_map, we should unregister it. otherwise, it will be polled every
+                    # time, which may cause 100% cpu usage
+                    log.debug('unexpected read event from fd %r', fd)
+                    try:
+                        self._context.poller.unregister_readable(fd)
+                    except Exception:  # noqa
+                        pass
+
+            for fd in w:
+                if fd in combined_map:
+                    try:
+                        dispatcher = combined_map[fd]
+                        log.debug('write event caused by %r', dispatcher)
+                        dispatcher.handle_write_event()
+                        if not dispatcher.writable():
+                            self._context.poller.unregister_writable(fd)
+                    except ExitNow:
+                        raise
+                    except Exception:  # noqa
+                        combined_map[fd].handle_error()
+                else:
+                    log.debug('unexpected write event from fd %r', fd)
+                    try:
+                        self._context.poller.unregister_writable(fd)
+                    except Exception:  # noqa
+                        pass
+
+            for group in pgroups:
+                group.transition()
+
+            self._reap()
+            self._handle_signal()
+            self._tick()
+
+            if self._context.state < SupervisorStates.RUNNING:
+                self._ordered_stop_groups_phase_2()
+
+            if self._context.test:
+                break
+
+    def _tick(self, now: ta.Optional[float] = None) -> None:
+        """Send one or more 'tick' events when the timeslice related to the period for the event type rolls over"""
+
+        if now is None:
+            # now won't be None in unit tests
+            now = time.time()
+
+        for event in TICK_EVENTS:
+            period = event.period  # type: ignore
+
+            last_tick = self._ticks.get(period)
+            if last_tick is None:
+                # we just started up
+                last_tick = self._ticks[period] = timeslice(period, now)
+
+            this_tick = timeslice(period, now)
+            if this_tick != last_tick:
+                self._ticks[period] = this_tick
+                notify_event(event(this_tick, self))
+
+    def _reap(self, *, once: bool = False, depth: int = 0) -> None:
+        if depth >= 100:
+            return
+
+        pid, sts = self._context.waitpid()
+        if not pid:
+            return
+
+        process = self._context.pid_history.get(pid, None)
+        if process is None:
+            _, msg = decode_wait_status(check_not_none(sts))
+            log.info('reaped unknown pid %s (%s)', pid, msg)
+        else:
+            process.finish(check_not_none(sts))
+            del self._context.pid_history[pid]
+
+        if not once:
+            # keep reaping until no more kids to reap, but don't recurse infinitely
+            self._reap(once=False, depth=depth + 1)
+
+    def _handle_signal(self) -> None:
+        sig = self._context.get_signal()
+        if not sig:
+            return
+
+        if sig in (signal.SIGTERM, signal.SIGINT, signal.SIGQUIT):
+            log.warning('received %s indicating exit request', signame(sig))
+            self._context.set_state(SupervisorStates.SHUTDOWN)
+
+        elif sig == signal.SIGHUP:
+            if self._context.state == SupervisorStates.SHUTDOWN:
+                log.warning('ignored %s indicating restart request (shutdown in progress)', signame(sig))  # noqa
+            else:
+                log.warning('received %s indicating restart request', signame(sig))  # noqa
+                self._context.set_state(SupervisorStates.RESTARTING)
+
+        elif sig == signal.SIGCHLD:
+            log.debug('received %s indicating a child quit', signame(sig))
+
+        elif sig == signal.SIGUSR2:
+            log.info('received %s indicating log reopen request', signame(sig))
+            # self._context.reopen_logs()
+            for group in self._process_groups.values():
+                group.reopen_logs()
+
+        else:
+            log.debug('received %s indicating nothing', signame(sig))
+
+
+def timeslice(period, when):
+    return int(when - (when % period))
+
+
+def main(args=None, test=False):
+    configure_standard_logging('INFO')
+
+    # if we hup, restart by making a new Supervisor()
+    first = True
+    while True:
+        config = ServerConfig.new(
+            nodaemon=True,
+            groups=[
+                ProcessGroupConfig(
+                    name='default',
+                    processes=[
+                        ProcessConfig(
+                            name='sleep',
+                            command='sleep 600',
+                            stdout=ProcessConfig.Log(
+                                file='/dev/fd/1',
+                                maxbytes=0,
+                            ),
+                            redirect_stderr=True,
+                        ),
+                        ProcessConfig(
+                            name='ls',
+                            command='ls -al',
+                            stdout=ProcessConfig.Log(
+                                file='/dev/fd/1',
+                                maxbytes=0,
+                            ),
+                            redirect_stderr=True,
+                        ),
+                    ],
+                ),
+            ],
+        )
+
+        context = ServerContext(
+            config,
+        )
+
+        context.first = first
+        context.test = test
+        go(context)
+        # options.close_logger()
+        first = False
+        if test or (context.state < SupervisorStates.RESTARTING):
+            break
+
+
+def go(context):  # pragma: no cover
+    d = Supervisor(context)
+    try:
+        d.main()
+    except ExitNow:
+        pass
+
+
+if __name__ == '__main__':
+    main()