使用Zabbix监控ZooKeeper服务的健康状态

一 应用场景描述

在目前公司的业务中，没有太多使用ZooKeeper作为协同服务的场景。但是我们将使用Codis作为Redis的集群部署方案，Codis依赖ZooKeeper来存储配置信息。所以做好ZooKeeper的监控也很重要。

二 ZooKeeper监控要点

系统监控

内存使用量 ZooKeeper应当完全运行在内存中，不能使用到SWAP。Java Heap大小不能超过可用内存。

Swap使用量 使用Swap会降低ZooKeeper的性能，设置vm.swappiness = 0

网络带宽占用 如果发现ZooKeeper性能降低关注下网络带宽占用情况和丢包情况，通常情况下ZooKeeper是20%写入80%读入

磁盘使用量 ZooKeeper数据目录使用情况需要注意

磁盘I/O ZooKeeper的磁盘写入是异步的，所以不会存在很大的I/O请求，如果ZooKeeper和其他I/O密集型服务公用应该关注下磁盘I/O情况

ZooKeeper监控

zk_avg/min/max_latency 响应一个客户端请求的时间，建议这个时间大于10个Tick就报警

zk_outstanding_requests 排队请求的数量，当ZooKeeper超过了它的处理能力时，这个值会增大，建议设置报警阀值为10

zk_packets_received 接收到客户端请求的包数量

zk_packets_sent 发送给客户单的包数量，主要是响应和通知

zk_max_file_descriptor_count 最大允许打开的文件数，由ulimit控制

zk_open_file_descriptor_count 打开文件数量，当这个值大于允许值得85%时报警

Mode 运行的角色，如果没有加入集群就是standalone,加入集群式follower或者leader

zk_followers leader角色才会有这个输出,集合中follower的个数。正常的值应该是集合成员的数量减1

zk_pending_syncs leader角色才会有这个输出，pending syncs的数量

zk_znode_count znodes的数量

zk_watch_count watches的数量

Java Heap Size ZooKeeper Java进程的

#echo ruok|nc 127.0.0.1 2181
imok

#echo mntr|nc 127.0.0.1 2181
zk_version 3.4.6-1569965, built on 02/20/2014 09:09 GMT
zk_avg_latency 0
zk_max_latency 0
zk_min_latency 0
zk_packets_received 11
zk_packets_sent 10
zk_num_alive_connections 1
zk_outstanding_requests 0
zk_server_state leader
zk_znode_count 17159
zk_watch_count 0
zk_ephemerals_count 1
zk_approximate_data_size 6666471
zk_open_file_descriptor_count 29
zk_max_file_descriptor_count 102400
zk_followers 2
zk_synced_followers 2
zk_pending_syncs 0

#echo srvr|nc 127.0.0.1 2181
Zookeeper version: 3.4.6-1569965, built on 02/20/2014 09:09 GMT
Latency min/avg/max: 0/0/0
Received: 26
Sent: 25
Connections: 1
Outstanding: 0
Zxid: 0x500000000
Mode: leader
Node count: 17159

三 编写Zabbix监控ZooKeeper的脚本和配置文件

要让Zabbix收集到这些监控数据，有两种方法一种是每个监控项目通过zabbix agent单独获取，主动监控和被动监控都可以。还有一种方法就是将这些监控数据一次性使用zabbix_sender全部发送给zabbix。这里我们选择第二种方式。那么采用zabbix_sender一次性发送全部监控数据的脚本就不能像通过zabbix agent这样逐个获取监控项目来编写脚本。

首先想办法将监控项目汇集成一个字典，然后遍历这个字典，将字典中的key:value对通过zabbix_sender的-k和-o参数指定发送出去

echo mntr|nc 127.0.0.1 2181

这条命令可以使用Python的subprocess模块调用，也可以使用socket模块去访问2181端口然后发送命令获取数据，获取到mntr执行的数据后还需要将其转化成为字典数据

即需要将这种样式的数据

zk_version 3.4.6-1569965, built on 02/20/2014 09:09 GMT
zk_avg_latency 0
zk_max_latency 0
zk_min_latency 0
zk_packets_received 91
zk_packets_sent 90
zk_num_alive_connections 1
zk_outstanding_requests 0
zk_server_state follower
zk_znode_count 17159
zk_watch_count 0
zk_ephemerals_count 1
zk_approximate_data_size 6666471
zk_open_file_descriptor_count 27
zk_max_file_descriptor_count 102400

转换成为这样的数据

{'zk_followers': 2, 'zk_outstanding_requests': 0, 'zk_approximate_data_size': 6666471, 'zk_packets_sent': 2089, 'zk_pending_syncs': 0, 'zk_avg_latency': 0, 'zk_version': '3.4.6-1569965, built on 02/20/2014 09:09 GMT', 'zk_watch_count': 2, 'zk_packets_received': 2090, 'zk_open_file_descriptor_count': 30, 'zk_server_ruok': 'imok', 'zk_server_state': 'leader', 'zk_synced_followers': 2, 'zk_max_latency': 28, 'zk_num_alive_connections': 2, 'zk_min_latency': 0, 'zk_ephemerals_count': 1, 'zk_znode_count': 17159, 'zk_max_file_descriptor_count': 102400}

到最后需要使用zabbix_sender发送的数据格式这个样子的

zookeeper.status[zk_version]这是key的名称

zookeeper.status[zk_outstanding_requests]:0
zookeeper.status[zk_approximate_data_size]:6666471
zookeeper.status[zk_packets_sent]:48
zookeeper.status[zk_avg_latency]:0
zookeeper.status[zk_version]:3.4.6-1569965, built on 02/20/2014 09:09 GMT
zookeeper.status[zk_watch_count]:0
zookeeper.status[zk_packets_received]:49
zookeeper.status[zk_open_file_descriptor_count]:27
zookeeper.status[zk_server_ruok]:imok
zookeeper.status[zk_server_state]:follower
zookeeper.status[zk_max_latency]:0
zookeeper.status[zk_num_alive_connections]:1
zookeeper.status[zk_min_latency]:0
zookeeper.status[zk_ephemerals_count]:1
zookeeper.status[zk_znode_count]:17159
zookeeper.status[zk_max_file_descriptor_count]:102400

精简代码如下：

#!/usr/bin/python
import socket
#from StringIO import StringIO
from cStringIO import StringIO
s=socket.socket()
s.connect(('localhost',2181))
s.send('mntr')
data_mntr=s.recv(2048)
s.close()
#print data_mntr
h=StringIO(data_mntr)
result={}
zresult={}
for line in h.readlines():
key,value=map(str.strip,line.split('\t'))
zkey='zookeeper.status' + '[' + key + ']'
zvalue=value
result[key]=value
zresult[zkey]=zvalue
print result
print '\n\n'
print zresult
#python test.py
{'zk_outstanding_requests': '0', 'zk_approximate_data_size': '6666471', 'zk_max_latency': '0', 'zk_avg_latency': '0', 'zk_version': '3.4.6-1569965, built on 02/20/2014 09:09 GMT', 'zk_watch_count': '0', 'zk_num_alive_connections': '1', 'zk_open_file_descriptor_count': '27', 'zk_server_state': 'follower', 'zk_packets_sent': '542', 'zk_packets_received': '543', 'zk_min_latency': '0', 'zk_ephemerals_count': '1', 'zk_znode_count': '17159', 'zk_max_file_descriptor_count': '102400'}

{'zookeeper.status[zk_watch_count]': '0', 'zookeeper.status[zk_avg_latency]': '0', 'zookeeper.status[zk_max_latency]': '0', 'zookeeper.status[zk_approximate_data_size]': '6666471', 'zookeeper.status[zk_server_state]': 'follower', 'zookeeper.status[zk_num_alive_connections]': '1', 'zookeeper.status[zk_min_latency]': '0', 'zookeeper.status[zk_outstanding_requests]': '0', 'zookeeper.status[zk_packets_received]': '543', 'zookeeper.status[zk_ephemerals_count]': '1', 'zookeeper.status[zk_znode_count]': '17159', 'zookeeper.status[zk_packets_sent]': '542', 'zookeeper.status[zk_open_file_descriptor_count]': '27', 'zookeeper.status[zk_max_file_descriptor_count]': '102400', 'zookeeper.status[zk_version]': '3.4.6-1569965, built on 02/20/2014 09:09 GMT'}

详细代码如下：

#!/usr/bin/python

""" Check Zookeeper Cluster

zookeeper version should be newer than 3.4.x

#echo mntr|nc 127.0.0.1 2181
zk_version 3.4.6-1569965, built on 02/20/2014 09:09 GMT
zk_avg_latency 0
zk_max_latency 4
zk_min_latency 0
zk_packets_received 84467
zk_packets_sent 84466
zk_num_alive_connections 3
zk_outstanding_requests 0
zk_server_state follower
zk_znode_count 17159
zk_watch_count 2
zk_ephemerals_count 1
zk_approximate_data_size 6666471
zk_open_file_descriptor_count 29
zk_max_file_descriptor_count 102400

#echo ruok|nc 127.0.0.1 2181
imok

"""

import sys
import socket
import re
import subprocess
from StringIO import StringIO
import os

zabbix_sender = '/opt/app/zabbix/sbin/zabbix_sender'
zabbix_conf = '/opt/app/zabbix/conf/zabbix_agentd.conf'
send_to_zabbix = 1

############# get zookeeper server status
class ZooKeeperServer(object):

def __init__(self, host='localhost', port='2181', timeout=1):
    self._address = (host, int(port))
    self._timeout = timeout
    self._result  = {}
 
def _create_socket(self):
    return socket.socket()
 
def _send_cmd(self, cmd):
    """ Send a 4letter word command to the server """
    s = self._create_socket()
    s.settimeout(self._timeout)
 
    s.connect(self._address)
    s.send(cmd)
 
    data = s.recv(2048)
    s.close()
 
    return data
 
def get_stats(self):
    """ Get ZooKeeper server stats as a map """
    data_mntr = self._send_cmd('mntr')
    data_ruok = self._send_cmd('ruok')
    if data_mntr:
        result_mntr = self._parse(data_mntr)
    if data_ruok:
        result_ruok = self._parse_ruok(data_ruok)
 
    self._result = dict(result_mntr.items() + result_ruok.items())
 
    if not self._result.has_key('zk_followers') and not self._result.has_key('zk_synced_followers') and not self._result.has_key('zk_pending_syncs'):
 
       ##### the tree metrics only exposed on leader role zookeeper server, we just set the followers' to 0
       leader_only = {'zk_followers':0,'zk_synced_followers':0,'zk_pending_syncs':0}    
       self._result = dict(result_mntr.items() + result_ruok.items() + leader_only.items() )
 
    return self._result  
 
def _parse(self, data):
    """ Parse the output from the 'mntr' 4letter word command """
    h = StringIO(data)
 
    result = {}
    for line in h.readlines():
        try:
            key, value = self._parse_line(line)
            result[key] = value
        except ValueError:
            pass # ignore broken lines
 
    return result
 
def _parse_ruok(self, data):
    """ Parse the output from the 'ruok' 4letter word command """
 
    h = StringIO(data)
 
    result = {}
 
    ruok = h.readline()
    if ruok:
       result['zk_server_ruok'] = ruok
 
    return result
 
def _parse_line(self, line):
    try:
        key, value = map(str.strip, line.split('\t'))
    except ValueError:
        raise ValueError('Found invalid line: %s' % line)
 
    if not key:
        raise ValueError('The key is mandatory and should not be empty')
 
    try:
        value = int(value)
    except (TypeError, ValueError):
        pass
 
    return key, value
 
def get_pid(self):

#ps -ef|grep java|grep zookeeper|awk '{print $2}'
pidarg = '''ps -ef|grep java|grep zookeeper|grep -v grep|awk '{print $2}' '''
pidout = subprocess.Popen(pidarg,shell=True,stdout=subprocess.PIPE)
pid = pidout.stdout.readline().strip('\n')
return pid

def send_to_zabbix(self, metric):
     key = "zookeeper.status[" +  metric + "]"
 
     if send_to_zabbix > 0:
         #print key + ":" + str(self._result[metric])
         try:
 
            subprocess.call([zabbix_sender, "-c", zabbix_conf, "-k", key, "-o", str(self._result[metric]) ], stdout=FNULL, stderr=FNULL, shell=False)
         except OSError, detail:
            print "Something went wrong while exectuting zabbix_sender : ", detail
     else:
            print "Simulation: the following command would be execucted :\n", zabbix_sender, "-c", zabbix_conf, "-k", key, "-o", self._result[metric], "\n"

def usage():
"""Display program usage"""

    print "\nUsage : ", sys.argv[0], " alive|all"
    print "Modes : \n\talive : Return pid of running zookeeper\n\tall : Send zookeeper stats as well"
    sys.exit(1)

accepted_modes = ['alive', 'all']

if len(sys.argv) == 2 and sys.argv[1] in accepted_modes:
mode = sys.argv[1]
else:
usage()

zk = ZooKeeperServer()
#print zk.get_stats()
pid = zk.get_pid()

if pid != "" and mode == 'all':
zk.get_stats()
#print zk._result
FNULL = open(os.devnull, 'w')
for key in zk._result:
zk.send_to_zabbix(key)
FNULL.close()
print pid

elif pid != "" and mode == "alive":
print pid
else:
print 0

zabbix配置文件check_zookeeper.conf

UserParameter=zookeeper.status[*],/usr/bin/python /opt/app/zabbix/sbin/check_zookeeper.py $1

重启agentd完成监控

此文章不作为你们参考，自己的做法还没有时间整理。后边在说。