/
usr
/
lib
/
python2.7
/
site-packages
/
fail2ban
/
server
/
Upload Filee
HOME
# emacs: -*- mode: python; py-indent-offset: 4; indent-tabs-mode: t -*- # vi: set ft=python sts=4 ts=4 sw=4 noet : # This file is part of Fail2Ban. # # Fail2Ban is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # # Fail2Ban is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with Fail2Ban; if not, write to the Free Software # Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. # Author: Cyril Jaquier # __author__ = "Cyril Jaquier" __copyright__ = "Copyright (c) 2004 Cyril Jaquier" __license__ = "GPL" import threading from threading import Lock, RLock import logging import os import signal import stat import sys from .observer import Observers, ObserverThread from .jails import Jails from .filter import FileFilter, JournalFilter from .transmitter import Transmitter from .asyncserver import AsyncServer, AsyncServerException from .. import version from ..helpers import getLogger, _as_bool, extractOptions, str2LogLevel, \ getVerbosityFormat, excepthook, prctl_set_th_name # Gets the instance of the logger. logSys = getLogger(__name__) DEF_SYSLOGSOCKET = "auto" DEF_LOGLEVEL = "INFO" DEF_LOGTARGET = "STDOUT" try: from .database import Fail2BanDb except ImportError: # pragma: no cover # Dont print error here, as database may not even be used Fail2BanDb = None def _thread_name(): return threading.current_thread().__class__.__name__ try: FileExistsError except NameError: # pragma: 3.x no cover FileExistsError = OSError def _make_file_path(name): """Creates path of file (last level only) on demand""" name = os.path.dirname(name) # only if it is absolute (e. g. important for socket, so if unix path): if os.path.isabs(name): # be sure path exists (create last level of directory on demand): try: os.mkdir(name) except (OSError, FileExistsError) as e: if e.errno != 17: # pragma: no cover - not EEXIST is not covered raise class Server: def __init__(self, daemon=False): self.__loggingLock = Lock() self.__lock = RLock() self.__jails = Jails() self.__db = None self.__daemon = daemon self.__transm = Transmitter(self) self.__reload_state = {} #self.__asyncServer = AsyncServer(self.__transm) self.__asyncServer = None self.__logLevel = None self.__logTarget = None self.__verbose = None self.__syslogSocket = None self.__autoSyslogSocketPaths = { 'Darwin': '/var/run/syslog', 'FreeBSD': '/var/run/log', 'Linux': '/dev/log', } self.__prev_signals = {} def __sigTERMhandler(self, signum, frame): # pragma: no cover - indirect tested logSys.debug("Caught signal %d. Exiting", signum) self.quit() def __sigUSR1handler(self, signum, fname): # pragma: no cover - indirect tested logSys.debug("Caught signal %d. Flushing logs", signum) self.flushLogs() def _rebindSignal(self, s, new): """Bind new signal handler while storing old one in _prev_signals""" self.__prev_signals[s] = signal.getsignal(s) signal.signal(s, new) def start(self, sock, pidfile, force=False, observer=True, conf={}): # First set the mask to only allow access to owner os.umask(0o077) # Second daemonize before logging etc, because it will close all handles: if self.__daemon: # pragma: no cover logSys.info("Starting in daemon mode") ret = self.__createDaemon() # If forked parent - return here (parent process will configure server later): if ret is None: return False # If error: if not ret[0]: err = "Could not create daemon %s", ret[1:] logSys.error(err) raise ServerInitializationError(err) # We are daemon. # replace main thread (and process) name to identify server (for top/ps/pstree or diagnostic): prctl_set_th_name(conf.get("pname", "fail2ban-server")) # Set all logging parameters (or use default if not specified): self.__verbose = conf.get("verbose", None) self.setSyslogSocket(conf.get("syslogsocket", self.__syslogSocket if self.__syslogSocket is not None else DEF_SYSLOGSOCKET)) self.setLogLevel(conf.get("loglevel", self.__logLevel if self.__logLevel is not None else DEF_LOGLEVEL)) self.setLogTarget(conf.get("logtarget", self.__logTarget if self.__logTarget is not None else DEF_LOGTARGET)) logSys.info("-"*50) logSys.info("Starting Fail2ban v%s", version.version) if self.__daemon: # pragma: no cover logSys.info("Daemon started") # Install signal handlers if _thread_name() == '_MainThread': for s in (signal.SIGTERM, signal.SIGINT): self._rebindSignal(s, self.__sigTERMhandler) self._rebindSignal(signal.SIGUSR1, self.__sigUSR1handler) # Ensure unhandled exceptions are logged sys.excepthook = excepthook # Creates a PID file. try: logSys.debug("Creating PID file %s", pidfile) _make_file_path(pidfile) pidFile = open(pidfile, 'w') pidFile.write("%s\n" % os.getpid()) pidFile.close() except (OSError, IOError) as e: # pragma: no cover logSys.error("Unable to create PID file: %s", e) # Create observers and start it: if observer: if Observers.Main is None: Observers.Main = ObserverThread() Observers.Main.start() # Start the communication logSys.debug("Starting communication") try: _make_file_path(sock) self.__asyncServer = AsyncServer(self.__transm) self.__asyncServer.onstart = conf.get('onstart') self.__asyncServer.start(sock, force) except AsyncServerException as e: logSys.error("Could not start server: %s", e) # Stop (if not yet already executed): self.quit() # Removes the PID file. try: logSys.debug("Remove PID file %s", pidfile) os.remove(pidfile) except (OSError, IOError) as e: # pragma: no cover logSys.error("Unable to remove PID file: %s", e) def quit(self): # Prevent to call quit twice: self.quit = lambda: False logSys.info("Shutdown in progress...") # Stop communication first because if jail's unban action # tries to communicate via fail2ban-client we get a lockup # among threads. So the simplest resolution is to stop all # communications first (which should be ok anyways since we # are exiting) # See https://github.com/fail2ban/fail2ban/issues/7 if self.__asyncServer is not None: self.__asyncServer.stop_communication() # Restore default signal handlers: if _thread_name() == '_MainThread': for s, sh in self.__prev_signals.iteritems(): signal.signal(s, sh) # Give observer a small chance to complete its work before exit obsMain = Observers.Main if obsMain is not None: if obsMain.stop(forceQuit=False): obsMain = None Observers.Main = None # Now stop all the jails self.stopAllJail() # Stop observer ultimately if obsMain is not None: obsMain.stop() # Explicit close database (server can leave in a thread, # so delayed GC can prevent commiting changes) if self.__db: self.__db.close() self.__db = None # Stop async and exit if self.__asyncServer is not None: self.__asyncServer.stop() self.__asyncServer = None logSys.info("Exiting Fail2ban") def addJail(self, name, backend): addflg = True if self.__reload_state.get(name) and self.__jails.exists(name): jail = self.__jails[name] # if backend switch - restart instead of reload: if jail.backend == backend: addflg = False logSys.info("Reload jail %r", name) # prevent to reload the same jail twice (temporary keep it in state, needed to commit reload): self.__reload_state[name] = None else: logSys.info("Restart jail %r (reason: %r != %r)", name, jail.backend, backend) self.delJail(name, stop=True) # prevent to start the same jail twice (no reload more - restart): del self.__reload_state[name] if addflg: self.__jails.add(name, backend, self.__db) if self.__db is not None: self.__db.addJail(self.__jails[name]) def delJail(self, name, stop=True, join=True): jail = self.__jails[name] if join or jail.isAlive(): jail.stop(stop=stop, join=join) if join: if self.__db is not None: self.__db.delJail(jail) del self.__jails[name] def startJail(self, name): with self.__lock: jail = self.__jails[name] if not jail.isAlive(): jail.start() elif name in self.__reload_state: logSys.info("Jail %r reloaded", name) del self.__reload_state[name] if jail.idle: jail.idle = False def stopJail(self, name): with self.__lock: self.delJail(name, stop=True) def stopAllJail(self): logSys.info("Stopping all jails") with self.__lock: # 1st stop all jails (signal and stop actions/filter thread): for name in self.__jails.keys(): self.delJail(name, stop=True, join=False) # 2nd wait for end and delete jails: for name in self.__jails.keys(): self.delJail(name, stop=False, join=True) def reloadJails(self, name, opts, begin): if begin: # begin reload: if self.__reload_state and (name == '--all' or self.__reload_state.get(name)): # pragma: no cover raise ValueError('Reload already in progress') logSys.info("Reload " + (("jail %s" % name) if name != '--all' else "all jails")) with self.__lock: # if single jail: if name != '--all': jail = None # test jail exists (throws exception if not): if "--if-exists" not in opts or self.__jails.exists(name): jail = self.__jails[name] if jail: # first unban all ips (will be not restored after (re)start): if "--unban" in opts: self.setUnbanIP(name) # stop if expected: if "--restart" in opts: self.stopJail(name) else: # first unban all ips (will be not restored after (re)start): if "--unban" in opts: self.setUnbanIP() # stop if expected: if "--restart" in opts: self.stopAllJail() # first set all affected jail(s) to idle and reset filter regex and other lists/dicts: for jn, jail in self.__jails.iteritems(): if name == '--all' or jn == name: jail.idle = True self.__reload_state[jn] = jail jail.filter.reload(begin=True) jail.actions.reload(begin=True) pass else: # end reload, all affected (or new) jails have already all new parameters (via stream) and (re)started: with self.__lock: deljails = [] for jn, jail in self.__jails.iteritems(): # still in reload state: if jn in self.__reload_state: # remove jails that are not reloaded (untouched, so not in new configuration) deljails.append(jn) else: # commit (reload was finished): jail.filter.reload(begin=False) jail.actions.reload(begin=False) for jn in deljails: self.delJail(jn) self.__reload_state = {} logSys.info("Reload finished.") def setIdleJail(self, name, value): self.__jails[name].idle = value return True def getIdleJail(self, name): return self.__jails[name].idle # Filter def setIgnoreSelf(self, name, value): self.__jails[name].filter.ignoreSelf = _as_bool(value) def getIgnoreSelf(self, name): return self.__jails[name].filter.ignoreSelf def addIgnoreIP(self, name, ip): self.__jails[name].filter.addIgnoreIP(ip) def delIgnoreIP(self, name, ip): self.__jails[name].filter.delIgnoreIP(ip) def getIgnoreIP(self, name): return self.__jails[name].filter.getIgnoreIP() def addLogPath(self, name, fileName, tail=False): filter_ = self.__jails[name].filter if isinstance(filter_, FileFilter): filter_.addLogPath(fileName, tail) def delLogPath(self, name, fileName): filter_ = self.__jails[name].filter if isinstance(filter_, FileFilter): filter_.delLogPath(fileName) def getLogPath(self, name): filter_ = self.__jails[name].filter if isinstance(filter_, FileFilter): return filter_.getLogPaths() else: # pragma: systemd no cover logSys.info("Jail %s is not a FileFilter instance" % name) return [] def addJournalMatch(self, name, match): # pragma: systemd no cover filter_ = self.__jails[name].filter if isinstance(filter_, JournalFilter): filter_.addJournalMatch(match) def delJournalMatch(self, name, match): # pragma: systemd no cover filter_ = self.__jails[name].filter if isinstance(filter_, JournalFilter): filter_.delJournalMatch(match) def getJournalMatch(self, name): # pragma: systemd no cover filter_ = self.__jails[name].filter if isinstance(filter_, JournalFilter): return filter_.getJournalMatch() else: logSys.info("Jail %s is not a JournalFilter instance" % name) return [] def setLogEncoding(self, name, encoding): filter_ = self.__jails[name].filter filter_.setLogEncoding(encoding) def getLogEncoding(self, name): filter_ = self.__jails[name].filter return filter_.getLogEncoding() def setFindTime(self, name, value): self.__jails[name].filter.setFindTime(value) def getFindTime(self, name): return self.__jails[name].filter.getFindTime() def setDatePattern(self, name, pattern): self.__jails[name].filter.setDatePattern(pattern) def getDatePattern(self, name): return self.__jails[name].filter.getDatePattern() def setLogTimeZone(self, name, tz): self.__jails[name].filter.setLogTimeZone(tz) def getLogTimeZone(self, name): return self.__jails[name].filter.getLogTimeZone() def setIgnoreCommand(self, name, value): self.__jails[name].filter.ignoreCommand = value def getIgnoreCommand(self, name): return self.__jails[name].filter.ignoreCommand def setIgnoreCache(self, name, value): value, options = extractOptions("cache["+value+"]") self.__jails[name].filter.ignoreCache = options def getIgnoreCache(self, name): return self.__jails[name].filter.ignoreCache def setPrefRegex(self, name, value): flt = self.__jails[name].filter logSys.debug(" prefregex: %r", value) flt.prefRegex = value def getPrefRegex(self, name): return self.__jails[name].filter.prefRegex def addFailRegex(self, name, value, multiple=False): flt = self.__jails[name].filter if not multiple: value = (value,) for value in value: logSys.debug(" failregex: %r", value) flt.addFailRegex(value) def delFailRegex(self, name, index=None): self.__jails[name].filter.delFailRegex(index) def getFailRegex(self, name): return self.__jails[name].filter.getFailRegex() def addIgnoreRegex(self, name, value, multiple=False): flt = self.__jails[name].filter if not multiple: value = (value,) for value in value: logSys.debug(" ignoreregex: %r", value) flt.addIgnoreRegex(value) def delIgnoreRegex(self, name, index): self.__jails[name].filter.delIgnoreRegex(index) def getIgnoreRegex(self, name): return self.__jails[name].filter.getIgnoreRegex() def setUseDns(self, name, value): self.__jails[name].filter.setUseDns(value) def getUseDns(self, name): return self.__jails[name].filter.getUseDns() def setMaxMatches(self, name, value): self.__jails[name].filter.failManager.maxMatches = value def getMaxMatches(self, name): return self.__jails[name].filter.failManager.maxMatches def setMaxRetry(self, name, value): self.__jails[name].filter.setMaxRetry(value) def getMaxRetry(self, name): return self.__jails[name].filter.getMaxRetry() def setMaxLines(self, name, value): self.__jails[name].filter.setMaxLines(value) def getMaxLines(self, name): return self.__jails[name].filter.getMaxLines() # Action def addAction(self, name, value, *args): ## create (or reload) jail action: self.__jails[name].actions.add(value, *args, reload=name in self.__reload_state) def getActions(self, name): return self.__jails[name].actions def delAction(self, name, value): del self.__jails[name].actions[value] def getAction(self, name, value): return self.__jails[name].actions[value] def setBanTime(self, name, value): self.__jails[name].actions.setBanTime(value) def addAttemptIP(self, name, *args): return self.__jails[name].filter.addAttempt(*args) def setBanIP(self, name, value): return self.__jails[name].actions.addBannedIP(value) def setUnbanIP(self, name=None, value=None, ifexists=True): if name is not None: # single jail: jails = [self.__jails[name]] else: # in all jails: jails = self.__jails.values() # unban given or all (if value is None): cnt = 0 ifexists |= (name is None) for jail in jails: cnt += jail.actions.removeBannedIP(value, ifexists=ifexists) return cnt def banned(self, name=None, ids=None): if name is not None: # single jail: jails = [self.__jails[name]] else: # in all jails: jails = self.__jails.values() # check banned ids: res = [] if name is None and ids: for ip in ids: ret = [] for jail in jails: if jail.actions.getBanned([ip]): ret.append(jail.name) res.append(ret) else: for jail in jails: ret = jail.actions.getBanned(ids) if name is not None: return ret res.append(ret) else: res.append({jail.name: ret}) return res def getBanTime(self, name): return self.__jails[name].actions.getBanTime() def getBanList(self, name, withTime=False): """Returns the list of banned IP addresses for a jail. Parameters ---------- name : str The name of a jail. Returns ------- list The list of banned IP addresses. """ return self.__jails[name].actions.getBanList(withTime) def setBanTimeExtra(self, name, opt, value): self.__jails[name].setBanTimeExtra(opt, value) def getBanTimeExtra(self, name, opt): return self.__jails[name].getBanTimeExtra(opt) def isStarted(self): return self.__asyncServer is not None and self.__asyncServer.isActive() def isAlive(self, jailnum=None): if jailnum is not None and len(self.__jails) != jailnum: return 0 for jail in self.__jails.values(): if not jail.isAlive(): return 0 return 1 # Status def status(self): try: self.__lock.acquire() jails = list(self.__jails) jails.sort() jailList = ", ".join(jails) ret = [("Number of jail", len(self.__jails)), ("Jail list", jailList)] return ret finally: self.__lock.release() def statusJail(self, name, flavor="basic"): return self.__jails[name].status(flavor=flavor) # Logging ## # Set the logging level. # # CRITICAL # ERROR # WARNING # NOTICE # INFO # DEBUG # @param value the level def setLogLevel(self, value): value = value.upper() with self.__loggingLock: if self.__logLevel == value: return ll = str2LogLevel(value) # don't change real log-level if running from the test cases: getLogger("fail2ban").setLevel( ll if DEF_LOGTARGET != "INHERITED" or ll < logging.DEBUG else DEF_LOGLEVEL) self.__logLevel = value ## # Get the logging level. # # @see setLogLevel # @return the log level def getLogLevel(self): with self.__loggingLock: return self.__logLevel ## # Sets the logging target. # # target can be a file, SYSLOG, STDOUT or STDERR. # @param target the logging target def setLogTarget(self, target): # check reserved targets in uppercase, don't change target, because it can be file: target, logOptions = extractOptions(target) systarget = target.upper() with self.__loggingLock: # don't set new handlers if already the same # or if "INHERITED" (foreground worker of the test cases, to prevent stop logging): if self.__logTarget == target: return True if systarget == "INHERITED": self.__logTarget = target return True padding = logOptions.get('padding') # set a format which is simpler for console use if systarget == "SYSLOG": facility = logOptions.get('facility', 'DAEMON').upper() # backwards compatibility - default no padding for syslog handler: if padding is None: padding = '0' try: facility = getattr(logging.handlers.SysLogHandler, 'LOG_' + facility) except AttributeError: # pragma: no cover logSys.error("Unable to set facility %r, using 'DAEMON'", logOptions.get('facility')) facility = logging.handlers.SysLogHandler.LOG_DAEMON if self.__syslogSocket == "auto": import platform self.__syslogSocket = self.__autoSyslogSocketPaths.get( platform.system()) if self.__syslogSocket is not None\ and os.path.exists(self.__syslogSocket)\ and stat.S_ISSOCK(os.stat( self.__syslogSocket).st_mode): hdlr = logging.handlers.SysLogHandler( self.__syslogSocket, facility=facility) else: logSys.error( "Syslog socket file: %s does not exists" " or is not a socket" % self.__syslogSocket) return False elif systarget in ("STDOUT", "SYSOUT"): hdlr = logging.StreamHandler(sys.stdout) elif systarget == "STDERR": hdlr = logging.StreamHandler(sys.stderr) else: # Target should be a file try: open(target, "a").close() hdlr = logging.handlers.RotatingFileHandler(target) except IOError: logSys.error("Unable to log to %r", target) logSys.info("Logging to previous target %r", self.__logTarget) return False # Removes previous handlers -- in reverse order since removeHandler # alter the list in-place and that can confuses the iterable logger = getLogger("fail2ban") for handler in logger.handlers[::-1]: # Remove the handler. logger.removeHandler(handler) # And try to close -- it might be closed already try: handler.flush() handler.close() except (ValueError, KeyError): # pragma: no cover # Is known to be thrown after logging was shutdown once # with older Pythons -- seems to be safe to ignore there # At least it was still failing on 2.6.2-0ubuntu1 (jaunty) if (2, 6, 3) <= sys.version_info < (3,) or \ (3, 2) <= sys.version_info: raise # detailed format by deep log levels (as DEBUG=10): if logger.getEffectiveLevel() <= logging.DEBUG: # pragma: no cover if self.__verbose is None: self.__verbose = logging.DEBUG - logger.getEffectiveLevel() + 1 # If handler don't already add date to the message: addtime = logOptions.get('datetime') if addtime is not None: addtime = _as_bool(addtime) else: addtime = systarget not in ("SYSLOG", "SYSOUT") if padding is not None: padding = _as_bool(padding) else: padding = True # If log-format is redefined in options: if logOptions.get('format', '') != '': fmt = logOptions.get('format') else: # verbose log-format: verbose = 0 if self.__verbose is not None and self.__verbose > 2: # pragma: no cover verbose = self.__verbose-1 fmt = getVerbosityFormat(verbose, addtime=addtime, padding=padding) # tell the handler to use this format hdlr.setFormatter(logging.Formatter(fmt)) logger.addHandler(hdlr) # Does not display this message at startup. if self.__logTarget is not None: logSys.info("Start Fail2ban v%s", version.version) logSys.info( "Changed logging target to %s for Fail2ban v%s" % ((target if target != "SYSLOG" else "%s (%s)" % (target, self.__syslogSocket)), version.version)) # Sets the logging target. self.__logTarget = target return True ## # Sets the syslog socket. # # syslogsocket is the full path to the syslog socket # @param syslogsocket the syslog socket path def setSyslogSocket(self, syslogsocket): with self.__loggingLock: if self.__syslogSocket == syslogsocket: return True self.__syslogSocket = syslogsocket # Conditionally reload, logtarget depends on socket path when SYSLOG return self.__logTarget != "SYSLOG"\ or self.setLogTarget(self.__logTarget) def getLogTarget(self): with self.__loggingLock: return self.__logTarget def getSyslogSocket(self): with self.__loggingLock: return self.__syslogSocket def flushLogs(self): if self.__logTarget not in ['STDERR', 'STDOUT', 'SYSLOG']: for handler in getLogger("fail2ban").handlers: try: handler.doRollover() logSys.info("rollover performed on %s" % self.__logTarget) except AttributeError: handler.flush() logSys.info("flush performed on %s" % self.__logTarget) return "rolled over" else: for handler in getLogger("fail2ban").handlers: handler.flush() logSys.info("flush performed on %s" % self.__logTarget) return "flushed" def setThreadOptions(self, value): for o, v in value.iteritems(): if o == 'stacksize': threading.stack_size(int(v)*1024) else: # pragma: no cover raise KeyError("unknown option %r" % o) def getThreadOptions(self): return {'stacksize': threading.stack_size() // 1024} def setDatabase(self, filename): # if not changed - nothing to do if self.__db and self.__db.filename == filename: return if not self.__db and filename.lower() == 'none': return if len(self.__jails) != 0: raise RuntimeError( "Cannot change database when there are jails present") if filename.lower() == "none": self.__db = None else: if Fail2BanDb is not None: _make_file_path(filename) self.__db = Fail2BanDb(filename) self.__db.delAllJails() else: # pragma: no cover logSys.error( "Unable to import fail2ban database module as sqlite " "is not available.") if Observers.Main is not None: Observers.Main.db_set(self.__db) def getDatabase(self): return self.__db def __createDaemon(self): # pragma: no cover """ Detach a process from the controlling terminal and run it in the background as a daemon. http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/278731 """ # When the first child terminates, all processes in the second child # are sent a SIGHUP, so it's ignored. # We need to set this in the parent process, so it gets inherited by the # child process, and this makes sure that it is effect even if the parent # terminates quickly. self._rebindSignal(signal.SIGHUP, signal.SIG_IGN) try: # Fork a child process so the parent can exit. This will return control # to the command line or shell. This is required so that the new process # is guaranteed not to be a process group leader. We have this guarantee # because the process GID of the parent is inherited by the child, but # the child gets a new PID, making it impossible for its PID to equal its # PGID. pid = os.fork() except OSError as e: return (False, (e.errno, e.strerror)) # ERROR (return a tuple) if pid == 0: # The first child. # Next we call os.setsid() to become the session leader of this new # session. The process also becomes the process group leader of the # new process group. Since a controlling terminal is associated with a # session, and this new session has not yet acquired a controlling # terminal our process now has no controlling terminal. This shouldn't # fail, since we're guaranteed that the child is not a process group # leader. os.setsid() try: # Fork a second child to prevent zombies. Since the first child is # a session leader without a controlling terminal, it's possible for # it to acquire one by opening a terminal in the future. This second # fork guarantees that the child is no longer a session leader, thus # preventing the daemon from ever acquiring a controlling terminal. pid = os.fork() # Fork a second child. except OSError as e: return (False, (e.errno, e.strerror)) # ERROR (return a tuple) if (pid == 0): # The second child. # Ensure that the daemon doesn't keep any directory in use. Failure # to do this could make a filesystem unmountable. os.chdir("/") else: os._exit(0) # Exit parent (the first child) of the second child. else: # Signal to exit, parent of the first child. return None # Close all open files. Try the system configuration variable, SC_OPEN_MAX, # for the maximum number of open files to close. If it doesn't exist, use # the default value (configurable). try: maxfd = os.sysconf("SC_OPEN_MAX") except (AttributeError, ValueError): maxfd = 256 # default maximum # urandom should not be closed in Python 3.4.0. Fixed in 3.4.1 # http://bugs.python.org/issue21207 if sys.version_info[0:3] == (3, 4, 0): # pragma: no cover urandom_fd = os.open("/dev/urandom", os.O_RDONLY) for fd in range(0, maxfd): try: if not os.path.sameopenfile(urandom_fd, fd): os.close(fd) except OSError: # ERROR (ignore) pass os.close(urandom_fd) else: os.closerange(0, maxfd) # Redirect the standard file descriptors to /dev/null. os.open("/dev/null", os.O_RDONLY) # standard input (0) os.open("/dev/null", os.O_RDWR) # standard output (1) os.open("/dev/null", os.O_RDWR) # standard error (2) return (True,) class ServerInitializationError(Exception): pass