Python获取服务端口存活状态并报警

103次阅读

没有评论

共计 4497 个字符，预计需要花费 12 分钟才能阅读完成。

最近发现公司的测试环境中有个 Socket 服务的端口总是莫名其妙 Down 掉，但是服务却正常运行着，看样子是僵死了。

虽然是测试环境，但是也不能这样放着不管，于是连夜写了一个简单的监控脚本。因为服务器是 Windows 的，所以要用到 wmi 模块。逻辑如下：

1、用 wmi 模块获取系统中处于停止状态的服务，生成一个字典。

2、判断监控的服务是否存在于字典中，如果存在说明服务已经停止，那么将尝试启动服务，并发送报警邮件。

3、向本地的 Socket 服务端口发送一个 connect，如果捕获到异常将尝试重启服务，并发送报警邮件。

4、每次执行时脚本将会循环执行以上步骤三次，间隔 10 秒，以确保服务状态正常。

在运行的时候发现了一个问题，Python 使用 wmi 模块来对 Windows 系统进行操作的时候速度格外的慢，不知道有没有其他的代替方法，哪位如果有更好的方法可以指点一下。

源码如下：

#!/usr/bin/env python

import os
import wmi
import time
import socket
import base64
import smtplib
import logging
from email.mime.text import MIMEText


def get_stop_service(designation):
    """Get stopped service name and caption,
    Filtration 'designation' service whether there is 'Stopped'.

    :return: service state
    """
    c = wmi.WMI()
    ret = dict()
    for service in c.Win32_Service():
        state, caption = service.State, service.Caption
        if state == 'Stopped':
            t = ret.get(state, [])
            t.append(caption)
            ret[state] = t
    # If 'designation' service in the 'Stopped', return status is 'down'
    if designation in ret.get('Stopped'):
        logging.error('Service [%s] is down, try to restart the service. \r\n' % designation)
        return 'down'
    return True


def monitor(sname):
    """Send the machine IP port 20000 socket request,
    If capture the abnormal returns the string 'ex'.

    :return: string 'ex'
    """
    s = socket.socket()
    s.settimeout(3)  # timeout
    host = ('127.0.0.1', 20000)
    try:  # Try connection to the host
        s.connect(host)
    except socket.error as e:
        logging.warning('[%s] service connection failed: %s \r\n' % (sname, e))
        return 'ex'
    return True


def restart_service(rstname, conn, run):
    """First check whether the service is stopped,
    if stop, start the service directly.
    The check whether the zombies,
    if a zombie, then restart the service.

    :return: flag or True
    """
    flag = False
    try:
        # From get_stop_service() to obtain the return value, the return value
        if run == 'down':
            ret = os.system('sc start"%s"' % rstname)
            if ret != 0:
                raise Exception('[Errno %s]' % ret)
            flag = True
        elif conn == 'ex':
            retStop = os.system('sc stop"%s"' % rstname)
            retSart = os.system('sc start"%s"' % rstname)
            if retSart != 0:
                raise Exception('retStop [Status code %s] '
                                'retSart [Status code %s] ' % (retStop, retSart))
            flag = True
        else:
            logging.info('[%s] service running status to normal' % rstname)
            return True
    except Exception as e:
        logging.warning('[%s] service restart failed: %s \r\n' % (rstname, e))
        return flag


def send_mail(to_list, sub, contents):
    """Send alarm mail.

    :return: flag
    """
    mail_server = 'mail.stmp.com'  # STMP Server
    mail_user = 'YouAccount'  # Mail account
    mail_pass = base64.b64decode('Password')  # The encrypted password
    mail_postfix = 'smtp.com'  # Domain name

    me = 'Monitor alarm<%s@%s>' % (mail_user, mail_postfix)
    message = MIMEText(contents, _subtype='html', _charset='utf-8')

    message['Subject'] = sub
    message['From'] = me
    message['To'] = ';'.join(to_list)

    flag = False  # To determine whether a mail sent successfully
    try:
        s = smtplib.SMTP()
        s.connect(mail_server)
        s.login(mail_user, mail_pass)
        s.sendmail(me, to_list, message.as_string())
        s.close()
        flag = True
    except Exception, e:
        logging.warning('Send mail failed, exception: [%s]. \r\n' % e)

    return flag


def main(sname):
    """Parameter type in the name of the service need to monitor,
    perform functions defined in turn, and the return value is correct.
    After the program is running, will test three times,
    each time interval to 10 seconds.

    :return: retValue
    """
    retry = 3
    count = 0
    retValue = False  # Used return to the state of the socket
    while count < retry:
        ret = monitor(sname)
        if ret != 'ex':  # If socket connection is normaol, return retValue
            retValue = ret
            return retValue
        isDown = get_stop_service(sname)
        restart_service(rstname=sname, conn=ret, run=isDown)

        host = socket.gethostname()
        address = socket.gethostbyname(host)
        mailto_list = ['mail@smtp.com', ]  # Alarm contacts
        send_mail(mailto_list, 
                  'Alarm',
                  ' <h4>Level: <u>ERROR</u></br> Host name: %s</br>'
                  ' IP Address: %s</br>'
                  ' Service name:</h4> <h5>%s</h5>'
                  % (host, address, sname))
        count += 1
        time.sleep(10)
    else:
        logging.error('[%s] service try to restart more than three times \r\n' % sname)

    return retValue


if __name__ == '__main__':

    logging.basicConfig(level=logging.INFO,
                        format='%(asctime)s %(levelname)s %(message)s',
                        datefmt='%Y/%m/%d %H:%M:%S',
                        filename='D:\\logs\\Monitor.log',
                        filemode='ab')

    name = 'Service Name'
    response = main(name)
    if response:
        logging.info('The [%s] service connection is normal \r\n' % name)