前言:目前有个需求,需要对线上几台拥有5个tomcat的服务器添加监控,分别监控各个tomcat的gc、连接数、使用内存等,现成的模板是监控一台tomcat,完全不适用目前情况,5个tomcat,如果现在新建模板,那每个tomcat得添加十几个item,也就是要添加60多个 Ttem
,而且还要添加 Trigger
和 Graph
,也都是成倍的添加,如果这台服务器在增加tomcat,又得再加一次,做工重复且繁琐,于是想到能不能自动发现tomca的端口以及把不同tomcat的监控数据直接发给zabbx-server。
此文章多方百度和多次实验失败后完成的。
参考1:zabbix_sender提交item数据
参考2:zabbix自动发现并监控主机的TCP监听端口
一、发现端口和pid
说明:发现端口是为了区分程序;而发现pid则是便于用jstat
获取程序的状态数据;
直接上脚本:
> vim jstat.py
#!/usr/bin/env python
#coding=utf-8
'''
##
## 功能: 调用jstat获取JMX的各项指标
## 说明: 用于zabbix自动发现告警
## 版本: V1.0 2016-11-02
## 特性: 1. 线程功能,提高脚本执行速度
##
'''
import sys
import os
import commands
import subprocess
import json
import argparse
import socket
import threading
jstat_cmd = commands.getoutput("which jstat")
jstack_cmd = commands.getoutput("which jstack")
jvmport_cmd = "netstat -tpnl|grep -oP '(?<=:)\d+.*\d+(?=/java)'|awk '{print $1,$NF}'|grep -E '808[0-9]|8899|7074'"
hostname = socket.gethostname()
zbx_sender='/usr/local/zabbix/bin/zabbix_sender'
zbx_cfg='/usr/local/zabbix/etc/zabbix_agentd.conf'
zbx_tmp_file='/usr/local/zabbix/scripts/.zabbix_jmx_status'
jstat_dict = {
"S0":"Young.Space0.Percent",
"S1":"Young.Space1.Percent",
"E":"Eden.Space.Percent",
"O":"Old.Space.Percent",
"P":"Perm.Space.Percent",
"FGC":"Old.Gc.Count",
"FGCT":"Old.Gc.Time",
"YGC":"Young.Gc.Count",
"YGCT":"Young.Gc.Time",
"GCT":"Total.Gc.Time",
"PGCMN":"Perm.Gc.Min",
"PGCMX":"Perm.Gc.Max",
"PGC":"Perm.Gc.New",
"PC":"Perm.Gc.Cur",
"Tomcat.Thread":"Tomcat.Thread"
}
jmx_threads = []
def get_status(cmd,opts,pid):
value = commands.getoutput('sudo %s -%s %s' % (cmd,opts,pid)).strip().split('\n')
kv = []
for i in value[0].split(' '):
if i != '':
kv.append(i)
vv = []
for i in value[1].split(' '):
if i != '':
vv.append(i)
data = dict(zip(kv,vv))
return data
def get_thread(cmd,pid):
value = commands.getoutput('sudo %s %s|grep http|wc -l' % (cmd,pid))
data = {"Tomcat.Thread":value}
return data
def get_jmx(jport,jprocess):
'''
使用jstat获取Java的性能指标
'''
file_truncate() # 清空zabbix_data_tmp
gcutil_data = get_status(jstat_cmd,"gcutil",jprocess)
gccapacity_data = get_status(jstat_cmd,"gccapacity",jprocess)
thread_data = get_thread(jstack_cmd,jprocess)
data_dict = dict(gcutil_data.items()+gccapacity_data.items()+thread_data.items())
for jmxkey in data_dict.keys():
if jmxkey in jstat_dict.keys():
cur_key = jstat_dict[jmxkey]
zbx_data = "%s jstat[%s,%s] %s" %(hostname,jport,cur_key,data_dict[jmxkey])
with open(zbx_tmp_file,'a') as file_obj: file_obj.write(zbx_data + '\n')
def jvm_port_discovery():
output = subprocess.Popen(jvmport_cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
jvm_port_lists = output.stdout.readlines()
jvm_port_proce = []
for jvm_port_tmp in jvm_port_lists:
jvm_port_proce.append(jvm_port_tmp.split())
return jvm_port_proce
def file_truncate():
'''
用于清空zabbix_sender使用的临时文件
'''
with open(zbx_tmp_file,'w') as fn: fn.truncate()
def zbx_tmp_file_create():
'''
创建zabbix_sender发送的文件内容
'''
jvmport_list = jvm_port_discovery()
for jvm_tmp in jvmport_list:
jvmport = jvm_tmp[0]
jvmprocess = jvm_tmp[1]
th = threading.Thread(target=get_jmx,args=(jvmport,jvmprocess))
th.start()
jmx_threads.append(th)
def send_data_zabbix():
'''
调用zabbix_sender命令,将收集的key和value发送至zabbix server
'''
zbx_tmp_file_create()
for get_jmxdata in jmx_threads:
get_jmxdata.join()
zbx_sender_cmd = "%s -c %s -i %s" %(zbx_sender,zbx_cfg,zbx_tmp_file)
print zbx_sender_cmd
zbx_sender_status,zbx_sender_result = commands.getstatusoutput(zbx_sender_cmd)
#print zbx_sender_status
print zbx_sender_result
def zbx_discovery():
'''
用于zabbix自动发现JVM端口
'''
jvm_zabbix = []
jvmport_list = jvm_port_discovery()
for jvm_tmp in jvmport_list:
jvm_zabbix.append({'{#JPORT}' : jvm_tmp[0],
'{#JPROCESS}' : jvm_tmp[1],
})
return json.dumps({'data': jvm_zabbix}, sort_keys=True, indent=7,separators=(',', ':'))
def cmd_line_opts(arg=None):
class ParseHelpFormat(argparse.HelpFormatter):
def __init__(self, prog, indent_increment=5, max_help_position=50, width=200):
super(ParseHelpFormat, self).__init__(prog, indent_increment, max_help_position, width)
parse = argparse.ArgumentParser(description='Jmx监控"',
formatter_class=ParseHelpFormat)
parse.add_argument('--version', '-v', action='version', version="0.1", help='查看版本')
parse.add_argument('--jvmport', action='store_true', help='获取JVM端口')
parse.add_argument('--data', action='store_true', help='发送JMX指标数据至zabbix')
if arg:
return parse.parse_args(arg)
if not sys.argv[1:]:
return parse.parse_args(['-h'])
else:
return parse.parse_args()
if __name__ == '__main__':
opts = cmd_line_opts()
if opts.jvmport:
print zbx_discovery()
elif opts.data:
send_data_zabbix()
else:
cmd_line_opts(arg=['-h'])
查看下返回数据
> sudo python jstat.py --jvmport
{
"data":[
{
"{#JPORT}":"8082",
"{#JPROCESS}":"456"
},
{
"{#JPORT}":"8083",
"{#JPROCESS}":"11992"
},
{
"{#JPORT}":"8084",
"{#JPROCESS}":"7713"
},
{
"{#JPORT}":"7074",
"{#JPROCESS}":"11239"
},
{
"{#JPORT}":"8899",
"{#JPROCESS}":"14186"
}
]
}
二、新建 templates
避免麻烦,我直接导出 templates
<?xml version="1.0" encoding="UTF-8"?>
<zabbix_export>
<version>2.0</version>
<date>2016-11-02T08:11:35Z</date>
<groups>
<group>
<name>Template java</name>
</group>
</groups>
<templates>
<template>
<template>jmx-im-jstat</template>
<name>jmx-im-jstat</name>
<groups>
<group>
<name>Template java</name>
</group>
</groups>
<applications>
<application>
<name>JSTAT</name>
</application>
</applications>
<items>
<item>
<name>jmxdata</name>
<type>0</type>
<snmp_community/>
<multiplier>1</multiplier>
<snmp_oid/>
<key>jmxdata</key>
<delay>30</delay>
<history>90</history>
<trends>365</trends>
<status>0</status>
<value_type>3</value_type>
<allowed_hosts/>
<units/>
<delta>0</delta>
<snmpv3_contextname/>
<snmpv3_securityname/>
<snmpv3_securitylevel>0</snmpv3_securitylevel>
<snmpv3_authprotocol>0</snmpv3_authprotocol>
<snmpv3_authpassphrase/>
<snmpv3_privprotocol>0</snmpv3_privprotocol>
<snmpv3_privpassphrase/>
<formula>1</formula>
<delay_flex/>
<params/>
<ipmi_sensor/>
<data_type>0</data_type>
<authtype>0</authtype>
<username/>
<password/>
<publickey/>
<privatekey/>
<port/>
<description/>
<inventory_link>0</inventory_link>
<applications>
<application>
<name>JSTAT</name>
</application>
</applications>
<valuemap/>
<logtimefmt/>
</item>
</items>
<discovery_rules>
<discovery_rule>
<name>jmxport</name>
<type>0</type>
<snmp_community/>
<snmp_oid/>
<key>jmxport</key>
<delay>30</delay>
<status>0</status>
<allowed_hosts/>
<snmpv3_contextname/>
<snmpv3_securityname/>
<snmpv3_securitylevel>0</snmpv3_securitylevel>
<snmpv3_authprotocol>0</snmpv3_authprotocol>
<snmpv3_authpassphrase/>
<snmpv3_privprotocol>0</snmpv3_privprotocol>
<snmpv3_privpassphrase/>
<delay_flex/>
<params/>
<ipmi_sensor/>
<authtype>0</authtype>
<username/>
<password/>
<publickey/>
<privatekey/>
<port/>
<filter>:</filter>
<lifetime>1</lifetime>
<description/>
<item_prototypes>
<item_prototype>
<name>port:$1 $2</name>
<type>2</type>
<snmp_community/>
<multiplier>1</multiplier>
<snmp_oid/>
<key>jstat[{#JPORT},Young.Gc.Count]</key>
<delay>0</delay>
<history>90</history>
<trends>365</trends>
<status>0</status>
<value_type>3</value_type>
<allowed_hosts/>
<units/>
<delta>0</delta>
<snmpv3_contextname/>
<snmpv3_securityname/>
<snmpv3_securitylevel>0</snmpv3_securitylevel>
<snmpv3_authprotocol>0</snmpv3_authprotocol>
<snmpv3_authpassphrase/>
<snmpv3_privprotocol>0</snmpv3_privprotocol>
<snmpv3_privpassphrase/>
<formula>1</formula>
<delay_flex/>
<params/>
<ipmi_sensor/>
<data_type>0</data_type>
<authtype>0</authtype>
<username/>
<password/>
<publickey/>
<privatekey/>
<port/>
<description/>
<inventory_link>0</inventory_link>
<applications>
<application>
<name>JSTAT</name>
</application>
</applications>
<valuemap/>
<logtimefmt/>
</item_prototype>
<item_prototype>
<name>port:$1 $2</name>
<type>2</type>
<snmp_community/>
<multiplier>1</multiplier>
<snmp_oid/>
<key>jstat[{#JPORT},Old.Gc.Count]</key>
<delay>0</delay>
<history>90</history>
<trends>365</trends>
<status>0</status>
<value_type>3</value_type>
<allowed_hosts/>
<units/>
<delta>0</delta>
<snmpv3_contextname/>
<snmpv3_securityname/>
<snmpv3_securitylevel>0</snmpv3_securitylevel>
<snmpv3_authprotocol>0</snmpv3_authprotocol>
<snmpv3_authpassphrase/>
<snmpv3_privprotocol>0</snmpv3_privprotocol>
<snmpv3_privpassphrase/>
<formula>1</formula>
<delay_flex/>
<params/>
<ipmi_sensor/>
<data_type>0</data_type>
<authtype>0</authtype>
<username/>
<password/>
<publickey/>
<privatekey/>
<port/>
<description/>
<inventory_link>0</inventory_link>
<applications>
<application>
<name>JSTAT</name>
</application>
</applications>
<valuemap/>
<logtimefmt/>
</item_prototype>
<item_prototype>
<name>port:$1 $2</name>
<type>2</type>
<snmp_community/>
<multiplier>1</multiplier>
<snmp_oid/>
<key>jstat[{#JPORT},Tomcat.Thread]</key>
<delay>0</delay>
<history>90</history>
<trends>365</trends>
<status>0</status>
<value_type>3</value_type>
<allowed_hosts/>
<units/>
<delta>0</delta>
<snmpv3_contextname/>
<snmpv3_securityname/>
<snmpv3_securitylevel>0</snmpv3_securitylevel>
<snmpv3_authprotocol>0</snmpv3_authprotocol>
<snmpv3_authpassphrase/>
<snmpv3_privprotocol>0</snmpv3_privprotocol>
<snmpv3_privpassphrase/>
<formula>1</formula>
<delay_flex/>
<params/>
<ipmi_sensor/>
<data_type>0</data_type>
<authtype>0</authtype>
<username/>
<password/>
<publickey/>
<privatekey/>
<port/>
<description/>
<inventory_link>0</inventory_link>
<applications>
<application>
<name>JSTAT</name>
</application>
</applications>
<valuemap/>
<logtimefmt/>
</item_prototype>
</item_prototypes>
<trigger_prototypes>
<trigger_prototype>
<expression>{jmx-im-jstat:jstat[{#JPORT},Tomcat.Thread].last(0)}>500</expression>
<name>Tomcat [#JPORT] Thread is too high</name>
<url/>
<status>0</status>
<priority>0</priority>
<description/>
<type>0</type>
</trigger_prototype>
</trigger_prototypes>
<graph_prototypes>
<graph_prototype>
<name>port:{#JPORT} Tomcat.Thread</name>
<width>900</width>
<height>200</height>
<yaxismin>0.0000</yaxismin>
<yaxismax>100.0000</yaxismax>
<show_work_period>1</show_work_period>
<show_triggers>1</show_triggers>
<type>0</type>
<show_legend>1</show_legend>
<show_3d>0</show_3d>
<percent_left>0.0000</percent_left>
<percent_right>0.0000</percent_right>
<ymin_type_1>0</ymin_type_1>
<ymax_type_1>0</ymax_type_1>
<ymin_item_1>0</ymin_item_1>
<ymax_item_1>0</ymax_item_1>
<graph_items>
<graph_item>
<sortorder>0</sortorder>
<drawtype>0</drawtype>
<color>C80000</color>
<yaxisside>0</yaxisside>
<calc_fnc>2</calc_fnc>
<type>0</type>
<item>
<host>jmx-im-jstat</host>
<key>jstat[{#JPORT},Tomcat.Thread]</key>
</item>
</graph_item>
</graph_items>
</graph_prototype>
</graph_prototypes>
<host_prototypes/>
</discovery_rule>
</discovery_rules>
<macros/>
<templates/>
<screens/>
</template>
</templates>
</zabbix_export>
注:模板太长,我只取2个item 放置其中;
下面附上几张配置截图
appname send-data discovery-port item- 目前只配置了item,还没有添加报警和图形
三、创建数据,并用zabbix_sender发送到zabbix-server
- zabbix_agentd配置文件
> cat /usr/local/zabbix/etc/zabbix_agentd.conf.d/jstat.conf
UserParameter=jmxport,sudo /usr/bin/python /usr/local/zabbix/scripts/jstat.py --jvmport
UserParameter=jmxdata,sudo /usr/bin/python /usr/local/zabbix/scripts/jstat.py --data
- 执行上面的脚本
> sudo python jstat.py --data
/usr/local/zabbix/bin/zabbix_sender -c /usr/local/zabbix/etc/zabbix_agentd.conf -i /usr/local/zabbix/scripts/.zabbix_jmx_status
info from server: "processed: 45; failed: 0; total: 45; seconds spent: 0.000531"
sent: 45; skipped: 0; total: 45
看上面结果,成功发送45个数据,数据会先临时存到 /usr/local/zabbix/scripts/.zabbix_jmx_status,我们查看下数据内容
> cat /usr/local/zabbix/scripts/.zabbix_jmx_status
test-01 jstat[8083,Young.Gc.Time] 36.572
test-01 jstat[8083,Old.Gc.Time] 3.971
test-01 jstat[8083,Perm.Gc.New] 80384.0
test-01 jstat[8083,Total.Gc.Time] 40.543
...
test-01 jstat[7074,Old.Gc.Time] 1.734
test-01 jstat[7074,Perm.Gc.New] 55296.0
test-01 jstat[7074,Total.Gc.Time] 26.635
第一列表示主机名,这个与zabbix_agentd.conf的配置保持一致,而且 zabbix-web添加的主机名也要一样;
第二列表示 key,这个和 templates里定义的一致;
第三列是 数据;
四、查看数据
data可以看到数据已经在zabbix展示了,之后我们可以添加相应的 报警和图形,使这个 template更加完善。
~ 全文完
网友评论