2018-4-26 Python实现对服务进程的监控，告警给钉钉机器人及重启任务

 1 #coding: utf-8
 2 import psutil
 3 import requests
 4 import time
 5 import os
 6 import json
 7 
 8 monitor_name = set([‘httpd‘,‘cobblerd‘])  # 用户指定监控的服务进程名称
 9 
10 proc_dict = {}
11 proc_name = set()  # 系统检测的进程名称
12 monitor_map = {
13     ‘httpd‘: ‘systemctl restart httpd‘,
14     ‘cobblerd‘: ‘systemctl restart cobblerd‘  # 系统在进程down掉后，自动重启
15 }
16 
17 dingding_url = ‘https://oapi.dingtalk.com/robot/send?access_token=b5258c4335ed8ab792075013c965efcbf4f8940f92e7bd936cdc7842d3bf9405‘
18 # 钉钉机器人token使用参考文档：http://www.pc6.com/infoview/Article_108931.html
19 
20 while True:
21     for proc in psutil.process_iter(attrs=[‘pid‘,‘name‘]):
22         proc_dict[proc.info[‘pid‘]] = proc.info[‘name‘]
23         proc_name.add(proc.info[‘name‘])
24 
25     proc_stop = monitor_name - proc_name  # 通过集合的形式来找出停掉的进程名,前者有但是后者没有的
26 
27     if proc_stop: # 如果确实有监控的进程停掉了，那么我们需要告警以及自动重启功能
28         for p in proc_stop:
29             p_status = ‘停止‘
30             p_name = p
31             data = {
32                 "msgtype": "markdown",
33                 "markdown": {
34                     "title": "监控信息",
35                     "text": "### %s\n" % time.strftime("%Y-%m-%d %X") +
36                     "> #### 服务名：%s \n\n" % p_name +
37                     "> #### 状态：%s \n\n" % p_status +
38                     "> #### 正在尝试启动"
39                 },
40             }
41             headers = {‘Content-Type‘:‘application/json;charset=UTF-8‘}
42             send_data = json.dumps(data).encode(‘utf-8‘)
43             requests.post(url=dingding_url,data=send_data,headers=headers)
44 
45             os.system(monitor_map[p_name]) # 执行重启命令，然后判断是否重启成功
46             proc_set = set()
47             for proc_again in psutil.process_iter(attrs=[‘pid‘,‘name‘]):
48                 proc_set.add(proc_again.info[‘name‘])
49 
50             if p in proc_set: # 如果进程启动成功，p是以前停掉的进程，proc_set是已经重启过一次后的所有进程集合
51                 p_status = ‘成功‘
52                 p_name = p
53                 data = {
54                     "msgtype": "markdown",
55                     "markdown": {
56                         "title": "监控信息",
57                         "text": "### %s\n" % time.strftime("%Y-%m-%d %X") +
58                         "> #### 服务名：%s \n\n" % p_name +
59                         "> #### 状态：%s \n\n" % p_status +
60                         "> #### 已经启动成功，服务正在运行！"
61                     },
62                 }
63                 headers = {‘Content-Type‘:‘application/json;charset=UTF-8‘}
64                 send_data = json.dumps(data).encode(‘utf-8‘)
65                 requests.post(url=dingding_url,data=send_data,headers=headers)
66             else:
67                 p_status = ‘重启失败‘
68                 p_name = p
69                 data = {
70                     "msgtype": "markdown",
71                     "markdown": {
72                         "title": "监控信息",
73                         "text": "### %s\n" % time.strftime("%Y-%m-%d %X") +
74                         "> #### 服务名：%s \n\n" % p_name +
75                         "> #### 状态：%s \n\n" % p_status +
76                         "> #### Sorry,服务启动失败鸟！"
77                     },
78                 }
79                 headers = {‘Content-Type‘:‘application/json;charset=UTF-8‘}
80                 send_data = json.dumps(data).encode(‘utf-8‘)
81                 requests.post(url=dingding_url,data=send_data,headers=headers)
82     time.sleep(5)