最近在写Daemon进程,在编写过程中遇到一些小麻烦,最终还是解决了。
我编写了两种,第一种是编写了一个程序,将其用setsid命令让其放入后台运行,第二种是直接fork()一个进程,在代码里将进程设置为后台启动。
在os.sytem()函数其他外部程序时,发现os.system()是阻塞的(os.popen()也是阻塞的),就是启动外部程序,你必须等外部程序退出,它才继续运行。用python中的subprocess库时,发现它并不阻塞主进程的运行,但是,你用外部kill命令杀死进程时,子进程会变成僵尸进程,只有父进程退出后才会退出。网上说在Windows平台下,python有个os.startfile是可以启动外部程序并不阻塞程序的运行,因为我写的Linux环境下,所以该函数不能用。最后问其他朋友,他说可以在system()将输出重定向就可以了,我试了一下,真的可以,所以现在把代码贴出了,也怪自己平常没有怎么钻。
os.system(processName+" 1>/dev/null 2>/dev/null &")
程序功能:
从配置文件读取要监控的进程,对进程实现监控,当监控程序退出时,会自动拉起进程
第一种方法:
在后台启动命令如下:setsid ./WatchProcessDog.py -m &
代码如下:
#!/usr/bin/python
#!encoding=utf-8
import ConfigParser
import sys
import threading
import time
import os
import commands
import subprocess
CONFIG_FILE = "WatchDog.ini"
SECTION="Monitor"
SECTION_KEY="Process"
class CWatchProcess(threading.Thread):
def __init__(self,configFile):
threading.Thread.__init__(self)
self.thread_stop = False
self.boolexist = False
self.curPid = os.getpid()
self.configFile = configFile
cfg = ConfigParser.ConfigParser()
try:
cfg.read(self.configFile)
allprocesses = cfg.get(SECTION,SECTION_KEY)
if '#' in allprocesses:
position1 = allprocesses.find('#')
self.processes = allprocesses[:position1]
else:
self.processes = allprocesses
self.processes = self.processes.strip()
self.monitorProcess = self.processes.split(',')
except Exception,e:
print e
def run(self):
while not self.thread_stop:
for tmpprocees in self.monitorProcess:
processname = os.path.basename(tmpprocees)
count = commands.getoutput("ps -elf | grep %s | grep -v %s | wc -l" % (processname,"grep"))
if 0 == int(count) and "-m" == sys.argv[1]:
self.PullProcess(tmpprocees)
continue
if 0 == int(count) and "-k" == sys.argv[1]:
self.boolexist = True
continue
if 0 != int(count) and "-m" == sys.argv[1]:
#print processname+" is started!"
continue
if 0 != int(count) and "-k" == sys.argv[1]:
self.KillProcess(tmpprocees)
self.boolexist = True
continue
if self.boolexist :
self.KillSelf()
time.sleep(1)
def stop(self):
self.thread_stop = True
def PullProcess(self,processName):
os.system(processName+" 1>/dev/null 2>/dev/null &")
def KillProcess(self,processName):
os.popen("killall %s" % os.path.basename(processName))
def KillSelf(self):
curNum = commands.getoutput("ps -elf | grep %s | grep -v %s | wc -l " % (sys.argv[0],"grep"))
if 1 == curNum:
sys.exit()
elif 1 < curNum:
result = os.popen("ps -elf | grep %s | grep -v %s" % (sys.argv[0],"grep")).readlines()
for tmpresult in result:
tmplist = tmpresult.split()
if int(self.curPid) == int(tmplist[3]):
continue
else:
os.popen("kill %s" % tmplist[3])
#tmpcount = len(tmplist)
sys.exit()
def main():
count = len(sys.argv)
if count != 2:
help()
sys.exit()
if "-m" == sys.argv[1]:
processNum = commands.getoutput("ps -elf | grep %s | grep -v %s | wc -l " % (sys.argv[0],"grep"))
if 1 < int(processNum):
print sys.argv[0],"is running"
sys.exit()
print "start monitor processes"
elif "-k" == sys.argv[1]:
print "kill all processes"
else:
help()
sys.exit()
def help():
print "Usage:"
print "./WatchProcessDog.py -m ---monitor all processes"
print "./WatchProcessDog.py -k ---kill all processes"
if __name__ == "__main__":
main()
watchProcess = CWatchProcess(CONFIG_FILE)
watchProcess.start()
watchProcess.join()
启动:./SqyDaemon.py -m
代码如下:
#!/usr/bin/env python
#!encoding=utf-8
import sys, os, time, atexit, string,ConfigParser,commands,subprocess
from signal import SIGTERM
PID_FILE = "./SqyDaemon.pid"
CONFIG_FILE = "SqyDaemon.ini"
SECTION="Monitor"
SECTION_KEY="Process"
class Daemon:
def __init__(self,configFile,pidfile):
self.pidfile = pidfile
self.configFile = configFile
cfg = ConfigParser.ConfigParser()
try:
cfg.read(self.configFile)
allprocesses = cfg.get(SECTION,SECTION_KEY)
if '#' in allprocesses:
position1 = allprocesses.find('#')
self.processes = allprocesses[:position1]
else:
self.processes = allprocesses
self.processes = self.processes.strip()
self.monitorProcess = self.processes.split(',')
except Exception,e:
print e
def _daemonize(self):
try:
pid = os.fork()
if pid > 0:
sys.exit(0) #退出主进程
except OSError, e:
print "fork failed!\nError is:",e.strerror
sys.exit(1)
os.setsid()
os.umask(0)
#创建子进程
try:
pid = os.fork()
if pid > 0:
sys.exit(0)
except OSError, e:
print "fork failed!\nError is:",e.strerror
sys.exit(1)
#创建processid文件
atexit.register(self.delpid)
pid = str(os.getpid())
file(self.pidfile,'w+').write('%s\n' % pid)
def delpid(self):
os.remove(self.pidfile)
def start(self):
#检查pid文件是否存在以探测是否存在进程
try:
pf = file(self.pidfile,'r')
pid = int(pf.read().strip())
pf.close()
except IOError:
pid = None
if pid:
print "pidfile %s already exist. SqyDaemon already running?\n" % self.pidfile
sys.exit(1)
#启动监控
self._daemonize()
self._run()
def stop(self):
#从pid文件中获取pid
try:
pf = file(self.pidfile,'r')
pid = int(pf.read().strip())
pf.close()
except IOError:
pid = None
if not pid:
if "-r" == sys.argv[1]:
print "SqyDaemon restart and monitor related process!"
else :
message = 'pidfile %s does not exist. SqyDaemon not running?\n'
sys.stderr.write(message % self.pidfile)
return #重启不报错
elif "-r" == sys.argv[1]:
print "%s is runing,now restart!" % sys.argv[0]
elif "-k" == sys.argv[1]:
print "all processes are killed!"
#杀进程
try:
while 1:
os.kill(pid, SIGTERM)
time.sleep(0.1)
for tmpprocees in self.monitorProcess:
processname = os.path.basename(tmpprocees)
os.system("killall %s" % processname)
except OSError, err:
err = str(err)
if err.find('No such process') > 0:
if os.path.exists(self.pidfile):
os.remove(self.pidfile)
else:
print str(err)
sys.exit(1)
def restart(self):
self.stop()
self.start()
def _run(self):
while True:
for tmpprocees in self.monitorProcess:
processname = os.path.basename(tmpprocees)
fullpath = os.path.abspath(tmpprocees)
count = commands.getoutput("ps -elf | grep %s | grep -v %s | wc -l" % (processname,"grep"))
if 0 == int(count):
os.system(tmpprocees+" 1>/dev/null 2>/dev/null &") #标准输出和错误输出重定向到/dev/null
else :
continue
time.sleep(2)
def help():
print "Usage:"
print "%s -m ---monitor all processes" % sys.argv[0]
print "%s -k ---kill all processes" % sys.argv[0]
print "%s -r ---restart all processes" % sys.argv[0]
if __name__ == '__main__':
daemon = Daemon(CONFIG_FILE,PID_FILE)
if len(sys.argv) == 2:
if '-m' == sys.argv[1]:
daemon.start()
elif '-k' == sys.argv[1]:
daemon.stop()
elif '-r' == sys.argv[1]:
daemon.restart()
else:
print 'Unknown command'
help()
sys.exit(2)
sys.exit(0)
else:
help()
sys.exit(2)