Daemonizing and logging added, 'stdin mode' fixed

tools/pypmmn/pypmmn/daemon.py

@@ -0,0 +1,210 @@
+## {{{ http://code.activestate.com/recipes/278731/ (r6)
+"""Disk And Execution MONitor (Daemon)
+
+Configurable daemon behaviors:
+
+   1.) The current working directory set to the "/" directory.
+   2.) The current file creation mode mask set to 0.
+   3.) Close all open files (1024). 
+   4.) Redirect standard I/O streams to "/dev/null".
+
+A failed call to fork() now raises an exception.
+
+References:
+   1) Advanced Programming in the Unix Environment: W. Richard Stevens
+   2) Unix Programming Frequently Asked Questions:
+         http://www.erlenstar.demon.co.uk/unix/faq_toc.html
+"""
+
+__author__ = "Chad J. Schroeder"
+__copyright__ = "Copyright (C) 2005 Chad J. Schroeder"
+
+__revision__ = "$Id$"
+__version__ = "0.2"
+
+# Standard Python modules.
+import os               # Miscellaneous OS interfaces.
+import sys              # System-specific parameters and functions.
+
+# Default daemon parameters.
+# File mode creation mask of the daemon.
+UMASK = 0
+
+# Default working directory for the daemon.
+WORKDIR = "/"
+
+# Default maximum for the number of available file descriptors.
+MAXFD = 1024
+
+# The standard I/O file descriptors are redirected to /dev/null by default.
+if (hasattr(os, "devnull")):
+   REDIRECT_TO = os.devnull
+else:
+   REDIRECT_TO = "/dev/null"
+
+def createDaemon():
+   """Detach a process from the controlling terminal and run it in the
+   background as a daemon.
+   """
+
+   try:
+      # Fork a child process so the parent can exit.  This returns control to
+      # the command-line or shell.  It also guarantees that the child will not
+      # be a process group leader, since the child receives a new process ID
+      # and inherits the parent's process group ID.  This step is required
+      # to insure that the next call to os.setsid is successful.
+      pid = os.fork()
+   except OSError, e:
+      raise Exception, "%s [%d]" % (e.strerror, e.errno)
+
+   if (pid == 0):   # The first child.
+      # To become the session leader of this new session and the process group
+      # leader of the new process group, we call os.setsid().  The process is
+      # also guaranteed not to have a controlling terminal.
+      os.setsid()
+
+      # Is ignoring SIGHUP necessary?
+      #
+      # It's often suggested that the SIGHUP signal should be ignored before
+      # the second fork to avoid premature termination of the process.  The
+      # reason is that when the first child terminates, all processes, e.g.
+      # the second child, in the orphaned group will be sent a SIGHUP.
+      #
+      # "However, as part of the session management system, there are exactly
+      # two cases where SIGHUP is sent on the death of a process:
+      #
+      #   1) When the process that dies is the session leader of a session that
+      #      is attached to a terminal device, SIGHUP is sent to all processes
+      #      in the foreground process group of that terminal device.
+      #   2) When the death of a process causes a process group to become
+      #      orphaned, and one or more processes in the orphaned group are
+      #      stopped, then SIGHUP and SIGCONT are sent to all members of the
+      #      orphaned group." [2]
+      #
+      # The first case can be ignored since the child is guaranteed not to have
+      # a controlling terminal.  The second case isn't so easy to dismiss.
+      # The process group is orphaned when the first child terminates and
+      # POSIX.1 requires that every STOPPED process in an orphaned process
+      # group be sent a SIGHUP signal followed by a SIGCONT signal.  Since the
+      # second child is not STOPPED though, we can safely forego ignoring the
+      # SIGHUP signal.  In any case, there are no ill-effects if it is ignored.
+      #
+      # import signal           # Set handlers for asynchronous events.
+      # signal.signal(signal.SIGHUP, signal.SIG_IGN)
+
+      try:
+         # Fork a second child and exit immediately to prevent zombies.  This
+         # causes the second child process to be orphaned, making the init
+         # process responsible for its cleanup.  And, 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 (System V-
+         # based systems).  This second fork guarantees that the child is no
+         # longer a session leader, preventing the daemon from ever acquiring
+         # a controlling terminal.
+         pid = os.fork()    # Fork a second child.
+      except OSError, e:
+         raise Exception, "%s [%d]" % (e.strerror, e.errno)
+
+      if (pid == 0):    # The second child.
+         # Since the current working directory may be a mounted filesystem, we
+         # avoid the issue of not being able to unmount the filesystem at
+         # shutdown time by changing it to the root directory.
+         os.chdir(WORKDIR)
+         # We probably don't want the file mode creation mask inherited from
+         # the parent, so we give the child complete control over permissions.
+         os.umask(UMASK)
+      else:
+         # exit() or _exit()?  See below.
+         os._exit(0)    # Exit parent (the first child) of the second child.
+   else:
+      # exit() or _exit()?
+      # _exit is like exit(), but it doesn't call any functions registered
+      # with atexit (and on_exit) or any registered signal handlers.  It also
+      # closes any open file descriptors.  Using exit() may cause all stdio
+      # streams to be flushed twice and any temporary files may be unexpectedly
+      # removed.  It's therefore recommended that child branches of a fork()
+      # and the parent branch(es) of a daemon use _exit().
+      os._exit(0)   # Exit parent of the first child.
+
+   # Close all open file descriptors.  This prevents the child from keeping
+   # open any file descriptors inherited from the parent.  There is a variety
+   # of methods to accomplish this task.  Three are listed below.
+   #
+   # Try the system configuration variable, SC_OPEN_MAX, to obtain the maximum
+   # number of open file descriptors to close.  If it doesn't exists, use
+   # the default value (configurable).
+   #
+   # try:
+   #    maxfd = os.sysconf("SC_OPEN_MAX")
+   # except (AttributeError, ValueError):
+   #    maxfd = MAXFD
+   #
+   # OR
+   #
+   # if (os.sysconf_names.has_key("SC_OPEN_MAX")):
+   #    maxfd = os.sysconf("SC_OPEN_MAX")
+   # else:
+   #    maxfd = MAXFD
+   #
+   # OR
+   #
+   # Use the getrlimit method to retrieve the maximum file descriptor number
+   # that can be opened by this process.  If there is not limit on the
+   # resource, use the default value.
+   #
+   import resource      # Resource usage information.
+   maxfd = resource.getrlimit(resource.RLIMIT_NOFILE)[1]
+   if (maxfd == resource.RLIM_INFINITY):
+      maxfd = MAXFD
+  
+   # Iterate through and close all file descriptors.
+   for fd in range(0, maxfd):
+      try:
+         os.close(fd)
+      except OSError:   # ERROR, fd wasn't open to begin with (ignored)
+         pass
+
+   # Redirect the standard I/O file descriptors to the specified file.  Since
+   # the daemon has no controlling terminal, most daemons redirect stdin,
+   # stdout, and stderr to /dev/null.  This is done to prevent side-effects
+   # from reads and writes to the standard I/O file descriptors.
+
+   # This call to open is guaranteed to return the lowest file descriptor,
+   # which will be 0 (stdin), since it was closed above.
+   os.open(REDIRECT_TO, os.O_RDWR)  # standard input (0)
+
+   # Duplicate standard input to standard output and standard error.
+   os.dup2(0, 1)            # standard output (1)
+   os.dup2(0, 2)            # standard error (2)
+
+   return(0)
+
+if __name__ == "__main__":
+
+   retCode = createDaemon()
+
+   # The code, as is, will create a new file in the root directory, when
+   # executed with superuser privileges.  The file will contain the following
+   # daemon related process parameters: return code, process ID, parent
+   # process group ID, session ID, user ID, effective user ID, real group ID,
+   # and the effective group ID.  Notice the relationship between the daemon's 
+   # process ID, process group ID, and its parent's process ID.
+
+   procParams = """
+   return code = %s
+   process ID = %s
+   parent process ID = %s
+   process group ID = %s
+   session ID = %s
+   user ID = %s
+   effective user ID = %s
+   real group ID = %s
+   effective group ID = %s
+   """ % (retCode, os.getpid(), os.getppid(), os.getpgrp(), os.getsid(0),
+   os.getuid(), os.geteuid(), os.getgid(), os.getegid())
+
+   open("createDaemon.log", "w").write(procParams + "\n")
+
+   sys.exit(retCode)
+## end of http://code.activestate.com/recipes/278731/ }}}
+

tools/pypmmn/pypmmn/pypmmn.py

@@ -1,20 +1,39 @@
 #!/usr/bin/python
+from logging.handlers import RotatingFileHandler
 from optparse import OptionParser
-from os import listdir, access, X_OK
+from os import listdir, access, X_OK, getpid
-from os.path import join, isdir
+from os.path import join, isdir, abspath, dirname, exists
 from subprocess import Popen, PIPE
-from socket import gethostname, SHUT_RDWR
 from time import sleep
+import logging
+import socket
 
 import sys
 
+LOG = logging.getLogger(__name__)
+LOG_FORMAT = '%(asctime)s - %(name)s - %(levelname)s - %(message)s'
 
-__version__ = '1.0dev3'
+from pypmmn.daemon import createDaemon
+
+
+__version__ = '1.0dev4'
 
 
 class CmdHandler(object):
+    """
+    This handler defines the protocol between munin and this munin node.
+    Each method starting with ``do_`` responds to the corresponding munin
+    command.
+    """
 
     def __init__(self, get_fun, put_fun, options):
+        """
+        Constructor
+
+        :param get_fun: The function used to receive a message from munin
+        :param put_fun: The function used to send a message back to munin
+        :param options: The command-line options object
+        """
         self.get_fun = get_fun
         self.put_fun = put_fun
         self.options = options
@@ -23,6 +42,7 @@ class CmdHandler(object):
         """
         Prints the version of this instance.
         """
+        LOG.debug('Command "version" executed with args: %r' % arg)
         self.put_fun('# munin node at %s\n' % (
             self.options.host,
             ))
@@ -31,26 +51,32 @@ class CmdHandler(object):
         """
         Prints this hostname
         """
+        LOG.debug('Command "nodes" executed with args: %r' % arg)
         self.put_fun('%s\n' % self.options.host)
-        self.put_fun('.')
+        self.put_fun('.\n')
 
     def do_quit(self, arg):
         """
         Stops this process
         """
+        LOG.debug('Command "quit" executed with args: %r' % arg)
         sys.exit(0)
 
     def do_list(self, arg):
         """
         Print a list of plugins
         """
+        LOG.debug('Command "list" executed with args: %r' % arg)
         try:
+            LOG.debug('Listing files inside %s' % self.options.plugin_dir)
             for filename in listdir(self.options.plugin_dir):
                 if not access(join(self.options.plugin_dir, filename), X_OK):
+                    LOG.warning('Non-executable plugin %s found!' % filename)
                     continue
+                #LOG.debug('Found plugin: %s' % filename)
                 self.put_fun("%s " % filename)
         except OSError, exc:
-            sys.stdout.write("# ERROR: %s" % exc)
+            self.put_fun("# ERROR: %s" % exc)
         self.put_fun("\n")
 
     def _caf(self, arg, cmd):
@@ -60,7 +86,9 @@ class CmdHandler(object):
         """
         plugin_filename = join(self.options.plugin_dir, arg)
         if isdir(plugin_filename) or not access(plugin_filename, X_OK):
-            self.put_fun("# Unknown plugin [%s] for %s" % (arg, cmd))
+            msg = "# Unknown plugin [%s] for %s" % (arg, cmd)
+            LOG.warning(msg)
+            self.put_fun(msg)
             return
 
         if cmd == 'fetch':
@@ -69,29 +97,55 @@ class CmdHandler(object):
             arg_plugin = cmd
 
         try:
-            output = Popen([plugin_filename, arg_plugin], stdout=PIPE).communicate()[0]
+            cmd = [plugin_filename, arg_plugin]
+            LOG.debug('Executing %r' % cmd)
+            output = Popen(cmd, stdout=PIPE).communicate()[0]
         except OSError, exc:
+            LOG.exception()
             self.put_fun("# ERROR: %s" % exc)
             return
         self.put_fun(output)
         self.put_fun('.\n')
 
     def do_alert(self, arg):
+        """
+        Handle command "alert"
+        """
+        LOG.debug('Command "alert" executed with args: %r' % arg)
         self._caf(arg, 'alert')
 
     def do_fetch(self, arg):
+        """
+        Handles command "fetch"
+        """
+        LOG.debug('Command "fetch" executed with args: %r' % arg)
         self._caf(arg, 'fetch')
 
     def do_config(self, arg):
+        """
+        Handles command "config"
+        """
+        LOG.debug('Command "config" executed with args: %r' % arg)
         self._caf(arg, 'config')
 
     def do_cap(self, arg):
+        """
+        Handles command "cap"
+        """
+        LOG.debug('Command "cap" executed with args: %r' % arg)
         self.put_fun("cap ")
         if self.options.spoolfetch_dir:
             self.put_fun("spool")
-        self.put_fun("cap \n")
+        else:
+            LOG.debug('No spoolfetch_dir specified. Result spooling disabled')
+
+        self.put_fun("\n")
 
     def do_spoolfetch(self, arg):
+        """
+        Handles command "spoolfetch"
+        """
+        LOG.debug('Command "spellfetch" executed with args: %r' % arg)
         output = Popen(['%s/spoolfetch_%s' % (self.options.spoolfetch_dir,
             self.options.host),
             arg]).communicate()[0]
@@ -101,8 +155,10 @@ class CmdHandler(object):
     # aliases
     do_exit = do_quit
 
-
     def handle_input(self, line):
+        """
+        Handles one input line and sends any result back using ``put_fun``
+        """
         line = line.strip()
         line = line.split(' ')
         cmd = line[0]
@@ -111,7 +167,8 @@ class CmdHandler(object):
         elif len(line) == 2:
             arg = line[1]
         else:
-            raise ValueError('Invalid input: %s' % line)
+            self.put_fun('# Invalid input: %s\n' % line)
+            return
 
         if not cmd:
             return
@@ -119,7 +176,8 @@ class CmdHandler(object):
         func = getattr(self, 'do_%s' % cmd, None)
         if not func:
             commands = [_[3:] for _ in dir(self) if _.startswith('do_')]
-            self.put_fun("# Unknown command. Supported commands: %s" % commands)
+            self.put_fun("# Unknown command. Supported commands: %s\n" % (
+                commands))
             return
 
         func(arg)
@@ -144,75 +202,169 @@ def get_options():
                 ' Default: <current working dir>/plugins'))
     parser.add_option('-h', '--host', dest='host',
             help=('The hostname which will be reported in the plugins.'
-                ' Default: %s' % gethostname()),
+                ' Default: %s' % socket.gethostname()),
-            default=gethostname())
+            default=socket.gethostname())
+    parser.add_option('-n', '--no-daemon', dest='no_daemon',
+            default=False,
+            action='store_true',
+            help='Run in foreground. Do not daemonize. '
+               'Will also enable debug logging to stdout.')
+    parser.add_option('-l', '--log-dir', dest='log_dir',
+            default=None,
+            help='The log folder. Default: disabled')
     parser.add_option('-s', '--spoolfech-dir', dest='spoolfetch_dir',
             default=None,
             help='The spoolfetch folder. Default: disabled')
     parser.add_option('--help', action='callback', callback=usage,
             help='Shows this help')
-    return parser.parse_args()
+
+    options, args = parser.parse_args()
+
+    # ensure we are using absolute paths (for daemonizing)
+    if options.log_dir:
+        options.log_dir = abspath(options.log_dir)
+
+    if options.spoolfetch_dir:
+        options.spoolfetch_dir = abspath(options.spoolfetch_dir)
+
+    if options.plugin_dir:
+        options.plugin_dir = abspath(options.plugin_dir)
+
+    return (options, args)
 
 
-def main():
+def process_stdin(options):
-    options, args = get_options()
+    rfhandler = RotatingFileHandler(
-    handler = CmdHandler(None, None, options)
+        join(abspath(dirname(__file__)), 'log', 'pypmmn.log'),
-    if not options.port:
+        maxBytes=100 * 1024,
-        handler.get_fun = sys.stdin.read
+        backupCount=5
-        handler.put_fun = sys.stdout.write
+        )
-    else:
+    rfhandler.setFormatter(logging.Formatter(LOG_FORMAT))
-        import socket
+    logging.getLogger().addHandler(rfhandler)
-        host = ''
+    handler = CmdHandler(sys.stdin.read, sys.stdout.write, options)
-        port = int(options.port)
+    handler.do_version(None)
-        s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+    LOG.info('STDIN handler opened')
-        s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+    while True:
-        s.bind((host, port))
+        data = sys.stdin.readline().strip()
-        s.listen(1)
+        handler.handle_input(data)
 
-        conn, addr = s.accept()
-        handler.get_fun = conn.recv
-        handler.put_fun = conn.send
-        handler.do_version(None)
-        counter = 0
 
-        print 'Connected by', addr
+def process_socket(options):
-        while True:
-            data = conn.recv(1024)
-            if not data.strip():
-                sleep(1)
-                counter += 1
-                if counter > 3:
-                    conn.shutdown(SHUT_RDWR)
-                    conn.close()
-                    conn, addr = s.accept()
-                    counter = 0
-                    handler.get_fun = conn.recv
-                    handler.put_fun = conn.send
-                    handler.do_version(None)
-                print "sleep"
-                try:
-                    data = conn.recv(1024)
-                    print 'data2', `data`
-                except socket.error, exc:
-                    conn, addr = s.accept()
-                    counter = 0
-                    handler.get_fun = conn.recv
-                    handler.put_fun = conn.send
-                    handler.do_version(None)
-                    print "Socket error: %s" % exc
 
-            if data.strip() == 'quit':
+    if options.no_daemon:
-                print 'shutting down remote connection'
+        # set up on-screen-logging
-                conn.shutdown(SHUT_RDWR)
+        console_handler = logging.StreamHandler(sys.stdout)
+        console_handler.setFormatter(logging.Formatter(LOG_FORMAT))
+        logging.getLogger().addHandler(console_handler)
+
+    retcode = 0
+    if not options.no_daemon:
+        retcode = createDaemon()
+        rfhandler = RotatingFileHandler(
+            join(options.log_dir, 'daemon.log'),
+            maxBytes=100 * 1024,
+            backupCount=5
+            )
+        rfhandler.setFormatter(logging.Formatter(LOG_FORMAT))
+        logging.getLogger().addHandler(rfhandler)
+        LOG.info('Process PID: %d' % getpid())
+        pidfile = open(join(options.log_dir, 'pypmmn.pid'), 'w')
+        pidfile.write(str(getpid()))
+        pidfile.close()
+        LOG.info('PID file created in %s' % join(options.log_dir,
+            'pypmmn.pid'))
+
+    LOG.info('Socket handler started.')
+
+    host = ''
+    port = int(options.port)
+    s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+    s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+    s.bind((host, port))
+    s.listen(1)
+
+    LOG.info('Listening on host %r, port %r' % (host, port))
+
+    conn, addr = s.accept()
+    handler = CmdHandler(conn.recv, conn.send, options)
+    handler.do_version(None)
+    counter = 0
+
+    LOG.info("Accepting incoming connection from %s" % (addr, ))
+    while True:
+        data = conn.recv(1024)
+        if not data.strip():
+            sleep(1)
+            counter += 1
+            if counter > 3:
+                LOG.info('Session timeout.')
+                conn.shutdown(socket.SHUT_RDWR)
                 conn.close()
+
+                LOG.info('Listening on host %r, port %r' % (host, port))
+
                 conn, addr = s.accept()
                 counter = 0
                 handler.get_fun = conn.recv
                 handler.put_fun = conn.send
                 handler.do_version(None)
-                continue
 
-            handler.handle_input(data)
+                LOG.info("Accepting incoming connection from %s" % (addr, ))
+            try:
+                data = conn.recv(1024)
+            except socket.error, exc:
+                LOG.warning("Socket error. Reinitialising.: %s" % exc)
+                conn, addr = s.accept()
+                counter = 0
+                handler.get_fun = conn.recv
+                handler.put_fun = conn.send
+                handler.do_version(None)
+
+                LOG.info("Accepting incoming connection from %s" % (addr, ))
+
+        if data.strip() == 'quit':
+            LOG.info('Client requested session end. Closing connection.')
+            conn.shutdown(socket.SHUT_RDWR)
+            conn.close()
+
+            LOG.info('Listening on host %r, port %r' % (host, port))
+
+            conn, addr = s.accept()
+            counter = 0
+            handler.get_fun = conn.recv
+            handler.put_fun = conn.send
+            handler.do_version(None)
+
+            LOG.info("Accepting incoming connection from %s" % (addr, ))
+
+            continue
+
+        handler.handle_input(data)
+
+    sys.exit(retcode)
+
+
+def main():
+    """
+    The main entry point of the application
+    """
+    options, args = get_options()
+
+    # Handle logging as early as possible.
+    if options.log_dir:
+        if not exists(options.log_dir):
+            raise IOError('[Errno 2] No such file or directory: %r' % (
+                options.log_dir))
+        # set up logging if requested
+        root_logger = logging.getLogger()
+        root_logger.setLevel(logging.NOTSET)  # TODO: Make configurable
+
+    # Start either the "stdin" interface, or the socked daemon. Depending on
+    # whether a port was given on startup or not.
+    if not options.port:
+        process_stdin(options)
+    else:
+        process_socket(options)
 
 
 if __name__ == '__main__':