zabbix监控docker
Centos6下安装easy_install# yum install python-setuptools
安装python 的docker模块
# easy_install docker-py
sudo:sorry, you must have a tty to run sudo
使用不同账户，执行执行脚本时候sudo经常会碰到 sudo: sorry, you must have a tty to run sudo这个情况，其实修改一下sudo的配置就好了
# vim /etc/sudoers (最好用visudo命令)
注释掉 Default requiretty 一行
#Default requiretty
意思就是sudo默认需要tty终端。注释掉就可以在后台执行了。
Zabbix客户端的部署：
#vim /opt/zabbix/etc/zabbix_agentd.conf
#docker
UserParameter=docker_discovery[*],cat/opt/zabbix/script/docker_cons.txt //用来发现宿主机上存活的容器
UserParameter=docker_stats[*],/opt/zabbix/script/zabbix_monitor_docker.py$1 $2 //用来监控容器的各种指标，后面会脚本具体体现，看不懂脚本的请路过。
UserParameter=docker.tomcat.discovery,cat/opt/zabbix/script/docker_tomcat.txt//用来发现容器启动的tomcat服务
UserParameter=docker.tomcat.stats[*],/opt/zabbix/script/zabbix_monitor_docker.py$1 $2 $3//用来监控容器中tomcat的端口
UserParameter=docker.nginx.discovery,cat/opt/zabbix/script/docker_nginx.txt//用来发现容器启动的nginx服务
UserParameter=docker.nginx.stats[*],/opt/zabbix/script/zabbix_monitor_docker.py$1 $2 $3 //用来监控容器中nginx的端口
监控脚本1，用来监控容器的CPU 内存 网卡，服务端口
#cat /opt/zabbix/script/zabbix_monitor_docker.py
#!/usr/bin/envpython#-*- coding:utf-8 -*-#email:279379936@qq.com
from dockerimport Clientimport sysimport subprocessimport osimport timeimport commands
defcheck_container_stats(container_name,collect_item): container_collect=docker_client.stats(container_name) container_collect.next() old_result=eval(container_collect.next()) new_result=eval(container_collect.next()) container_collect.close() if collect_item == 'cpu_total_usage': result=new_result['cpu_stats']['cpu_usage']['total_usage'] -old_result['cpu_stats']['cpu_usage']['total_usage'] elif collect_item == 'cpu_system_usage': result=new_result['cpu_stats']['system_cpu_usage'] -old_result['cpu_stats']['system_cpu_usage'] elif collect_item == 'cpu_percent': cpu_total_usage=new_result['cpu_stats']['cpu_usage']['total_usage'] -old_result['cpu_stats']['cpu_usage']['total_usage'] cpu_system_uasge=new_result['cpu_stats']['system_cpu_usage'] -old_result['cpu_stats']['system_cpu_usage'] cpu_num=len(old_result['cpu_stats']['cpu_usage']['percpu_usage']) result=round((float(cpu_total_usage)/float(cpu_system_uasge))*cpu_num*100.0,2) elif collect_item == 'mem_usage': result=new_result['memory_stats']['usage'] elif collect_item == 'mem_limit': result=new_result['memory_stats']['limit'] elif collect_item == 'mem_percent': mem_usage=new_result['memory_stats']['usage'] mem_limit=new_result['memory_stats']['limit'] result=round(float(mem_usage)/float(mem_limit)*100.0,2) elif collect_item == 'network_rx_bytes': network_check_command="""dockerexec %s ifconfig eth1|grep bytes|awk -F ':' '{print $2,$3}'|awk -F '(' '{print$1,$2}'|awk -F ')' '{print $1}'|awk '{print"{\\"rx\\":"$1",\\"tx\\":"$2"}"}'"""%container_name network_old_result=eval(((subprocess.Popen(network_check_command,shell=True,stdout=subprocess.PIPE)).stdout.readlines()).strip('\n')) #print time.time() #print network_old_result time.sleep(1) network_new_result=eval(((subprocess.Popen(network_check_command,shell=True,stdout=subprocess.PIPE)).stdout.readlines()).strip('\n')) #print time.time() #print network_new_result #unit b result=int(network_new_result['rx']) -int(network_old_result['rx']) elif collect_item == 'network_tx_bytes': network_check_command="""dockerexec %s ifconfig eth1|grep bytes|awk -F ':' '{print $2,$3}'|awk -F '(' '{print$1,$2}'|awk -F ')' '{print $1}'|awk '{print"{\\"rx\\":"$1",\\"tx\\":"$2"}"}'"""%container_name network_old_result=eval(((subprocess.Popen(network_check_command,shell=True,stdout=subprocess.PIPE)).stdout.readlines()).strip('\n')) time.sleep(1) network_new_result=eval(((subprocess.Popen(network_check_command,shell=True,stdout=subprocess.PIPE)).stdout.readlines()).strip('\n')) result=int(network_new_result['tx']) -int(network_old_result['tx']) return resultif __name__ =="__main__": docker_client = Client(base_url='unix://var/run/docker.sock',version='1.19') if len(sys.argv) == 3: container_name=sys.argv collect_item=sys.argv printcheck_container_stats(container_name,collect_item) elif len(sys.argv) == 4 and sys.argv =='port': container_name=sys.argv collect_item=int(sys.argv) check_stat=commands.getoutput("/usr/bin/docker exec %s netstat-ntpul|grep %s > /dev/null;echo $?" %(container_name,collect_item)) print check_stat else: print '1'说明：上面脚本为通过python的docker模块去抓取数据，由于各种原因，有些机器无法安装python模块，可通过下面脚本实现：# cat /opt/zabbix/script/zabbix_monitor_docker.py #!/usr/bin/envpython#-*- coding:utf-8 -*-#email:279379936@qq.com
import sysimportsubprocessimport timeimport commandsimport re
defget_memory_container_dir(memory_dir,container_name): con_id=commands.getoutput("sudo/usr/bin/docker ps|grep %s|awk '{print $1}'" % container_name) con_full_id=commands.getoutput("ls -al%s|grep '%s'|grep -v grep|awk '{print $NF}'" % (memory_dir,con_id)) memory_container_dir=memory_dir + '/' +con_full_id return memory_container_dir
defget_cpu_container_dir(cpu_dir,container_name): con_id=commands.getoutput("sudo/usr/bin/docker ps|grep %s|awk '{print $1}'" % container_name) con_full_id=commands.getoutput("ls -al%s|grep '%s'|grep -v grep|awk '{print $NF}'" % (cpu_dir,con_id)) cpu_container_dir=cpu_dir + '/' +con_full_id return cpu_container_dirdefget_cpu_info(container_name): info = commands.getoutput('echo -ne"GET /containers/%s/stats?stream=false HTTP/1.1\r\n\r\n"|sudo/usr/bin/nc -U /var/run/docker.sock|grep read' % container_name) info = eval(info) return info
defcheck_container_stats(container_name,collect_item): if collect_item == 'cpu_total_usage': old_result =get_cpu_info(container_name) new_result =get_cpu_info(container_name) old_time = old_result['read'] new_time = new_result['read'] list_old_time = re.split('\:|\.',old_time) list_new_time =re.split('\:|\.',new_time) old_s = int(list_old_time) new_s = int(list_new_time) if old_s >= new_s: time_interval = 2 else: time_interval = new_s - old_s result=(new_result['cpu_stats']['cpu_usage']['total_usage']- old_result['cpu_stats']['cpu_usage']['total_usage']) / time_interval elif collect_item == 'cpu_system_usage':
old_result =get_cpu_info(container_name) new_result = get_cpu_info(container_name) old_time = old_result['read'] new_time = new_result['read'] list_old_time =re.split('\:|\.',old_time) list_new_time =re.split('\:|\.',new_time) old_s = int(list_old_time) new_s = int(list_new_time) if old_s >= new_s: time_interval = 2 else: time_interval = new_s - old_s result=(new_result['cpu_stats']['system_cpu_usage'] -old_result['cpu_stats']['system_cpu_usage']) / time_interval elif collect_item == 'cpu_percent':
old_result =get_cpu_info(container_name) new_result = get_cpu_info(container_name) old_time = old_result['read'] new_time = new_result['read'] list_old_time =re.split('\:|\.',old_time) list_new_time =re.split('\:|\.',new_time) old_s = int(list_old_time) new_s = int(list_new_time) if old_s >= new_s: time_interval = 2 else: time_interval = new_s - old_s cpu_total_usage=(new_result['cpu_stats']['cpu_usage']['total_usage'] -old_result['cpu_stats']['cpu_usage']['total_usage']) / time_interval cpu_system_uasge=(new_result['cpu_stats']['system_cpu_usage'] -old_result['cpu_stats']['system_cpu_usage']) / time_interval cpu_num=len(old_result['cpu_stats']['cpu_usage']['percpu_usage']) result=round((float(cpu_total_usage)/float(cpu_system_uasge))*cpu_num*100.0,2) elif collect_item == 'mem_usage':
memory_container_dir=get_memory_container_dir(memory_dir,container_name) result=commands.getoutput("cat%s/memory.stat|grep '^rss'|grep -v grep|awk '{print $NF}'" %memory_container_dir) elif collect_item == 'mem_limit':
memory_container_dir=get_memory_container_dir(memory_dir,container_name) result=commands.getoutput("cat%s/memory.limit_in_bytes" % memory_container_dir) elif collect_item == 'mem_percent':
memory_container_dir=get_memory_container_dir(memory_dir,container_name) mem_usage=commands.getoutput("cat%s/memory.stat|grep '^rss'|grep -v grep|awk '{print $NF}'" %memory_container_dir) mem_limit=commands.getoutput("cat%s/memory.limit_in_bytes" % memory_container_dir) result=round(float(mem_usage)/float(mem_limit)*100.0,2) elif collect_item == 'network_rx_bytes':
network_check_command="""sudo /usr/bin/docker exec %s ifconfigeth1|grep bytes|awk -F':' '{print $2,$3}'|awk '{print $1,$6}'|awk '{print"{\\"rx\\":"$1",\\"tx\\":"$2"}"}'"""%container_name network_old_result=eval(((subprocess.Popen(network_check_command,shell=True,stdout=subprocess.PIPE)).stdout.readlines()).strip('\n')) time.sleep(1) network_new_result=eval(((subprocess.Popen(network_check_command,shell=True,stdout=subprocess.PIPE)).stdout.readlines()).strip('\n')) result=int(network_new_result['rx']) -int(network_old_result['rx']) elif collect_item == 'network_tx_bytes':
network_check_command="""sudo /usr/bin/docker exec %sifconfig eth1|grep bytes|awk -F':' '{print $2,$3}'|awk '{print $1,$6}'|awk'{print "{\\"rx\\":"$1",\\"tx\\":"$2"}"}'"""%container_name network_old_result=eval(((subprocess.Popen(network_check_command,shell=True,stdout=subprocess.PIPE)).stdout.readlines()).strip('\n')) time.sleep(1) network_new_result=eval(((subprocess.Popen(network_check_command,shell=True,stdout=subprocess.PIPE)).stdout.readlines()).strip('\n')) result=int(network_new_result['tx']) -int(network_old_result['tx']) return result
if __name__ =="__main__": cpu_dir="/cgroup/cpuacct/docker" memory_dir="/cgroup/memory/docker" iflen(sys.argv) == 3: container_name=sys.argv collect_item=sys.argv printcheck_container_stats(container_name,collect_item) elif len(sys.argv) == 4 and sys.argv =='port': container_name=sys.argv collect_item=int(sys.argv) check_stat=commands.getoutput("sudo /usr/bin/docker exec %s netstat-ntpul|grep %s > /dev/null;echo $?" %(container_name,collect_item)) print check_stat else: print '1'
脚本2，用来发现容器名
catdiscovery_cons.py
#!/usr/bin/env python
# Felix Shang
#QQ: 279379936
import commands
import sys
def docker_s():
cons = commands.getoutput("""sudo /usr/bin/docker ps|grep-v "CONTAINER ID"|awk '{print $NF}'|tr '\n' ' '""")
count_cons = len(cons.split())
if count_cons != 0:
return cons.split()
else:
return 0
if __name__ == "__main__":
if len(sys.argv) == 2 and sys.argv == 'docker':
infos = docker_s()
if infos != 0:
print '{'
print '\t"data":['
i = 0
cou_infos=len(infos)
for con in infos:
if i == cou_infos - 1:
print'\t\t{"{#CONTAINERNAME}":"%s"}' % con
else:
print'\t\t{"{#CONTAINERNAME}":"%s"},' % con
i = i + 1
print '\t]'
print '}'
脚本3，用来发现容器的服务（tomcat nginx），之前脚本2和脚本3是一个脚本，发现容器时出现好多问题。
# cat/opt/zabbix/script/discovery_docker_service.py
#!/usr/bin/env python
# Felix Shang
#QQ: 279379936
import commands
import sys
def docker_s():
cons = commands.getoutput("""cat /opt/zabbix/script/docker_cons.txt|grep'CONTAINERNAME'|grep -v grep|awk -F'"' '{print $4}'|tr '\n''\t'""")
#print cons
count_cons = len(cons.split())
if count_cons != 0:
return cons.split()
else:
return 0
def tomcat_s():
cons = docker_s()
if cons == 0:
sys.exit(2)
else:
cons_d = {}
for con in cons:
#print con
stat = commands.getoutput("sudo /usr/bin/docker exec %s ps -ef|grepjava|grep tomcat|grep -v grep>/dev/null;echo $?" % con)
port_list = []
if int(stat) == 0:
tomcat_config_dirs =commands.getoutput("sudo /usr/bin/docker exec %s ps -ef | grep tomcat |grep -v grep | awk -F\= '{print $2}' | awk -F'logging' '{print $1}'" %con).split()
for tomcat_config_dir intomcat_config_dirs:
tomcat_config_file =tomcat_config_dir + 'server.xml'
port =commands.getoutput("""sudo /usr/bin/docker exec %s grep"port=" %s|grep -v "shutdown"|grep -v "AJP"|grep"Connector"|awk -F\= '{print $2}'|awk '{print $1}'"""%(con,tomcat_config_file)).strip('"')
port_list.append(port)
cons_d = port_list
else:
cons_d = port_list
return cons_d
def nginx_s():
cons = docker_s()
if cons == 0:
sys.exit(2)
else:
cons_d = {}
for con in cons:
stat = commands.getoutput("sudo /usr/bin/docker exec %s ps -ef|grepnginx|grep -v grep>/dev/null;echo $?" % con)
port_list = []
if int(stat) == 0:
port_list =commands.getoutput("sudo /usr/bin/docker exec %s netstat -ntpul|grepnginx|grep -v 40080|awk '{print $4}'|awk -F\: '{print $NF}'|tr '\n' ' '" %con).split()
cons_d = port_list
else:
cons_d = port_list
return cons_d
if __name__ == "__main__":
if len(sys.argv) == 2 and sys.argv == 'tomcat':
infos = tomcat_s()
print '{'
print '\t"data":['
port_infos = []
for con_info in infos:
if len(infos) == 0:
continue
else:
for port in infos:
port_info ='\t\t{"{#CONTAINERNAME}":"%s","{#CON_TOMCAT_PORT}":"%s"},'%(con_info,port)
port_infos.append(port_info)
i = 0
cou_port_infos = len(port_infos)
for port_i in port_infos:
if i == cou_port_infos - 1:
port_i = port_i
print port_i
i = i + 1
print '\t]'
print '}'
elif len(sys.argv) == 2 and sys.argv == 'nginx':
infos = nginx_s()
print '{'
print '\t"data":['
port_infos = []
for con_info in infos:
if len(infos) == 0:
continue
else:
for port in infos:
port_info ='\t\t{"{#CONTAINERNAME}":"%s","{#CON_NGINX_PORT}":"%s"},'%(con_info,port)
port_infos.append(port_info)
i = 0
cou_port_infos = len(port_infos)
for port_i in port_infos:
if i == cou_port_infos - 1:
port_i = port_i
print port_i
i = i + 1
print '\t]'
print '}'
#else:
# help_s()
#vim /etc/sudoers //zabbix_agent是通过zabbix用户执行，通过sudo提权让zabbix用户对脚本有执行权限。
zabbix ALL=(root) NOPASSWD:/usr/bin/docker,/sbin/fdisk,/usr/sbin/dmidecode,/usr/bin/nc
Zabbix服务端的配置：
导入模板：Template docker, 宿主机关联此模板即可。
报错：
Server获取值报错：ZBX_NOTSUPPORTED]
# vim zabbix_agentd.conf
# 设置超时时间
Timeout=30
页:
[1]