zabbix问题事件统计/按月统计/生成表格/Python调用Zabbix3.4api

背景：

因业务需要，每月初需要为客户提供指定预警等级的事件，这里选择从evnet.get中进行获取。

发现的坑：

目前维护多套zabbix服务,并且版本不同..， POC进行时使用的是5.4.2, 正式环境3.4.x， 3.4版本中的event.get是没有name属性的，

最后我是通过trigger拿到的信息，但因为tigger中带着{HOST.NAME} 或 {HOST.CONN} 这块儿通过ip类的信息进行了正则替换，具体如代码，比较简单，主要是API使用思路。

目的：

获取事件名称、开始时间、主机名称和IP信息，整理后以表格展示

调取结构图：

代码环境：

Python3.8

代码依赖包：

pip install openpyxl

pip install pandas

pip install requests

代码：

  1 import requests
  2 import json
  3 import pandas as pd
  4 import time
  5 import re
  6 
  7 def stamp2time(timestamp): #时间戳转日期函数
  8     time_local = time.localtime(int(timestamp))
  9     dt = time.strftime("%Y-%m-%d %H:%M:%S", time_local)
 10     return dt
 11 
 12 class zabbix_api(object):
 13     def __init__(self):
 14         self.header = {"Content-Type": "application/json-rpc"}
 15         self.url = 'https://zabbix.xxx.com.cn/zabbix/api_jsonrpc.php'
 16         self.auth = self.zabbix_login()
 17 
 18     def zabbix_login(self):
 19         data = {
 20             "jsonrpc": "2.0",
 21             "method": "user.login",
 22             "params": {
 23                 "user": "xxx",  # zabbix登录用户名
 24                 "password": "xxxx"  # zabbix登录密码
 25             },
 26             "id": 1,
 27             "auth": None,
 28         }
 29         return self.requestJson(data)
 30 
 31     def getHostsip(self, hostid):
 32         data = {'jsonrpc': '2.0',
 33                   'method': 'host.get',
 34                   'params': {
 35                       'output': ["hostid", "host", "name"],
 36                       'selectInterfaces': ["ip"],
 37                       'hostids': hostid
 38                   },
 39                   'auth': self.auth,
 40                   'id': '10'
 41                   }
 42         output = self.requestJson(data)
 43         return {"hostname": output[0]['host'], "ip": output[0]['interfaces'][0]['ip']}
 44 
 45     def getHosts(self):
 46         hostsip = []
 47         data = host_data = {
 48             "jsonrpc": "2.0",
 49             "method": "host.get",
 50             "params": {},
 51             "auth": self.auth,
 52             "id": 1
 53         }
 54         output = self.requestJson(data)
 55         for hosts in output[0:]:
 56             hostsip.append(hosts['interfaces'])
 57         return hostsip
 58 
 59     def getEvents(self):
 60         hostsip = []
 61         data = {
 62             "jsonrpc": "2.0",
 63             "method": "event.get",
 64             "params": {
 65                 "output": ["clock"],  # 输出内容, 这里原始数据只需要时间
 66                 "time_from": "1625760000",  # 开始时间
 67                 "time_till": "1628438400",     # 结束时间
 68                 "selectHosts": "extend",         # 显示主机相关信息
 69                 "value": 1,  # 只显示问题事件，不显示恢复时间（默认：都显示）
 70                 # "select_alerts": "extend",  # 通过此参数获取次事件触发的报警信息
 71                 "selectRelatedObject": "extend",   # 通过此参数扩展获得对应触发器信息
 72                 "severities": [2, 3, 4, 5],    # 获取等级
 73                 "sortfield": ["clock", "eventid"],  # 排序属性
 74                 "sortorder": "desc"
 75             },
 76             "auth": self.auth,
 77             "id": 2
 78         }
 79         output = self.requestJson(data)
 80         return output
 81 
 82     def getTriggers(self, triggerid):
 83         data = {
 84             "jsonrpc": "2.0",
 85             "method": "trigger.get",
 86             "params": {
 87                 "triggerids": triggerid,
 88                 "output": [
 89                     "triggerid",
 90                     "description",
 91                     "name"
 92                 ],
 93                 "sortfield": "priority",
 94                 "sortorder": "DESC"
 95             },
 96             "auth": self.auth,
 97             "id": 1
 98         }
 99         output = self.requestJson(data)
100         return output
101 
102     def requestJson(self, json_values):
103         r = json.loads(requests.post(self.url, data=json.dumps(json_values), headers=self.header).content.decode())
104         return r["result"]
105 
106 
107 z = zabbix_api()
108 event_list = []
109 print(len(z.getEvents()))
110 data_lens = 0
111 for event in z.getEvents():
112     data_lens += 1
113     event_dict = {}
114     if event['hosts']:
115         # print(event)
116         hostid = event['hosts'][0]['hostid']
117         event_dict['clock'] = event['clock']
118         event_dict['hostname'] = z.getHostsip(hostid)['hostname']
119         event_dict['ip'] = z.getHostsip(hostid)['ip']
120         trigger_name = event['relatedObject']['description']
121         if re.findall("{HOST.NAME}", trigger_name):
122             event_dict['name'] = trigger_name.replace("{HOST.NAME}", z.getHostsip(hostid)['hostname'])
123         elif re.findall("{HOST.CONN}", trigger_name):
124             event_dict['name'] = trigger_name.replace("{HOST.CONN}", z.getHostsip(hostid)['ip'])
125         else:
126             event_dict['name'] = trigger_name
127         print(event_dict)
128         # event_dict['name'] = event['name']
129         event_list.append(event_dict)
130 
131 df = pd.DataFrame(event_list)
132 print(df)
133 df['clock'] = df['clock'].apply(stamp2time)
134 df.to_excel("result_xlsx.xlsx", index=0)

有待改进：

- 各个参数的传递

- 事件数量过多的话，整个统计过程会非常缓慢，需要进行多线程优化

参考：

- 获取zabbix 所有的hostid

https://www.jianshu.com/p/7a8a765b5992?utm_campaign=maleskine&utm_content=note&utm_medium=seo_notes&utm_source=recommendation

- zabbix3.4 event.get

https://www.zabbix.com/documentation/3.4/manual/api/reference/event/get

- zabbix3.4 host.get

https://www.zabbix.com/documentation/3.4/manual/api/reference/host/get

- pandas 时间戳与日期互转方法

https://zhuanlan.zhihu.com/p/138130809

- python写入数据到csv或xlsx文件的3种方法

https://www.jb51.net/article/168327.htm

- 在线时间戳转换工具

https://tool.lu/timestamp/