python监控服务
脚本
#!/usr/bin/python
# coding:utf-8
import socket, os, re, sys, sendmail, psutil, time
with open('/home/ctchat/python/error.log', 'a+') as f:
class Myserver():
fuwu_ip = str(re.findall(":(\d.*)\s{2}B", os.popen("ifconfig").read())[0])
port_lst = []
def __init__(self, service, number):
self.service = service
self.number = number
self.shell_service = "netstat -tnlp|grep " + " " + service
self.file_name_set1 = "/home/ctchat/python/" + self.service + "set1.txt"
self.file_name_set2 = "/home/ctchat/python/" + self.service + "set2.txt"
def run(self):
result = os.popen(self.shell_service).read()
return result
def Guo_lu(self):
post = list(set(re.findall(".*:(\d+)", self.run())))
return post
def jiance_all(self):
try:
res = 0
f.write(' --- ' + self.service + ':' + str(self.Guo_lu()))
if len(self.Guo_lu()) >= int(self.number):
for i in range(len(self.Guo_lu())):
if int(self.Guo_lu()[i]) in [1805, 1905, 8015, 19966]:
continue
else:
res += Myserver.get(self.fuwu_ip, int(self.Guo_lu()[i]))
f.write(' --- socket.port.num:' + str(res) + ' --- socket.port:' + str(Myserver.port_lst))
if res > 0:
return 1
else:
return 0
else:
f.write('\n********************\n'+self.run() + '\n********************\n')
return 0
except IndexError:
return 1
@staticmethod
def get(ip, port):
try:
Myserver.port_lst.append(port)
client = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
client.connect((ip, port))
return 1
except:
return 0
finally:
client.close()
class Process():
all = psutil.pids()
def __init__(self, name, number):
self.name = name
self.number = number
def process_num(self):
try:
n = 0
for i in range(len(Process.all)):
all_class = psutil.Process(Process.all[i])
if all_class.name() == self.name:
n += 1
f.write(str(time.strftime('[%Y-%m-%d %H:%M:%S]',
time.localtime(time.time()))) + ' --- ' + self.name + '.process.num:' + str(n))
if n >= int(self.number):
return 1
else:
return 0
except psutil.NoSuchProcess:
f.write(str(time.strftime('[%Y-%m-%d %H:%M:%S]',
time.localtime(time.time()))) + ' --- ' + self.name + ':psutil.NoSuchProcess' )
return 1
res1 = Process(sys.argv[3], sys.argv[4]).process_num()
res2 = Myserver(sys.argv[1], sys.argv[2]).jiance_all()
f.write(' --- port.result:'+ str(res2) +' --- porcess.result:'+ str(res1) + '\n')
if res1 == 1 and res2 == 1:
print(1)
else:
print(0)
agent端操作
[root@server02 ctchat]# mkdir /home/ctchat/python [root@server02 ctchat]# cd /home/ctchat/python
将脚本上传到服务器python
[root@server02 python]# chmod +x get_port.py [root@server02 python]# chmod 777 error.log # 返回1代表成功 [root@server02 python]# ./get_port.py mysqld 1 mysqld 1 1
查看当前目录生成有error.log文件，日志如果跑时间太长可以使用 echo > error.log 清空
[root@server02 python]# ls error.log get_port.py
查看内容
[root@server02 python]# cat error.log [2020-03-05 17:03:25] --- mysqld.process.num:1 --- mysqld:['1342'] --- socket.port.num:1 --- socket.port:[1342] --- port.result:1 --- porcess.result:1
解析：
- mysqld.process.num:1 # 进程数量为1个
- mysqld:['1342'] # 最主要的，脚本过滤出来的端口号
- socket.port:[1342] # 监测通过的端口号
- port.result:1 # 端口检查结果1为正常，当为0时告警
- porcess.result:1 # 进程监测的结果，同上
配置agent
[root@server02 python]# vi /etc/zabbix/zabbix_agentd.d/userparameter_service.conf UserParameter=service.status[*], /usr/bin/python /home/ctchat/python/get_port.py $1 $2 $3 $4 [root@server02 python]# /etc/init.d/zabbix-agent restart
zabbix_server执行
[root@server01 ~]# zabbix_get -s 10.4.7.6 -p 10050 -k "service.status[mysqld 1 mysqld 1]" (Not all processes could be identified, non-owned process info will not be shown, you would have to be root to see it all.) (Not all processes could be identified, non-owned process info will not be shown, you would have to be root to see it all.) (Not all processes could be identified, non-owned process info will not be shown, you would have to be root to see it all.) 0
如果出现上面情况，需要在新部署的agent端执行chmod +s /bin/netstat
[root@server01 ~]# zabbix_get -s 10.4.7.6 -p 10050 -k "service.status[mysqld 1 mysqld 1]" 1
页面配置
创建模板
创建应用集
固定格式，只需要更改服务名称
{mysql:service.status[mysqld 1 mysqld 1].last(0)}=0
last(0)代表收到为0的值之后立即报警
可以添加一个监控图看看是否配置成功
