一、Kafka监控的几个指标
1、lag:多少消息没有消费 lag=logsize-offset2、logsize:Kafka存的消息总数
3、offset:已经消费的消息
Kafka管理工具 介绍:
二、查看zookeeper配置
cat /home/app/zookeeper/zookeeper/conf/zoo.cfg | egrep -v "^$|^#"
clientPort=2181
三、查看kafka配置
cat /home/app/kafka/kafka/config/server.properties | egrep -v "^$|^#"
port=9092
host.name=192.168.38.87
zookeeper.connect=192.168.38.87:2181,192.168.38.88:2181
四、查看kafka的group name
cd /home/app/zookeeper/zookeeper/bin
./zkCli.sh -server 192.168.38.87:2181
ls /consumers/
lijieGroup
quit
五、查看kafka的topic_name
/home/app/kafka/kafka/bin/kafka-run-class.sh kafka.tools.ConsumerOffsetChecker --group=lijieGroup --zookeeper=192.168.38.87:2181
六、修改zabbix配置文件
因为zabbix用户不能调用kafka的脚本,需要root用户启动zabbix_agent
vim /opt/zabbix/zabbix_agentd.conf
AllowRoot=1
User=root
Include=/opt/zabbix/zabbix_agentd.conf.d/
vim /opt/zabbix/zabbix_agentd.conf.d/kafka_status.conf
UserParameter=kafka.lag[*],/home/zabbix_scripts/kafka_mon.sh $1 $2 lag
UserParameter=kafka.offset[*],/home/zabbix_scripts/kafka_mon.sh $1 $2 offset
UserParameter=kafka.logsize[*],/home/zabbix_scripts/kafka_mon.sh $1 $2 logsize
chown -R zabbix:zabbix /opt/zabbix/zabbix_agentd.conf.d/kafka_status.conf
chmod -R 777 /opt/zabbix/zabbix_agentd.conf.d/kafka_status.conf
七、创建监控脚本
mkdir -pv /home/zabbix_scripts/
vim /home/zabbix_scripts/kafka_mon.sh
#!/bin/bash
kafka_ip="127.0.0.1"
kafka_port=2181
topic_name=$1
group_id=$2
pn=$3
/usr/local/kafka_2.11-0.11.0.0/bin/kafka-run-class.sh kafka.tools.ConsumerOffsetChecker --topic=$topic_name --group=$group_id --zookeeper=$kafka_ip:$kafka_port | grep -v Offset > /tmp/kafka-tp-${topic_name}-${group_id}.info
Offset=0
logSize=0
Lag=0
while read line
do
Offset=$((${Offset}+`echo $line |awk ‘{print $4}‘`))
logSize=$((${logSize}+`echo $line |awk ‘{print $5}‘`))
Lag=$(($Lag+`echo $line |awk ‘{print $6}‘`))
done < /tmp/kafka_topic_info.log
case $pn in
offset|Offset)
echo $Offset
;;
logsize|logSize)
echo $logSize
;;
lag|Lag)
echo $Lag
;;
*)
echo Error
;;
esac
八、给脚本和对应文件权限
chown -R zabbix:zabbix /home/zabbix_scripts/kafka_mon.sh
chmod -R 777 /home/zabbix_scripts/kafka_mon.sh
touch /tmp/kafka-tp-RouterOnOfflineStateChange-lijieGroup.info
chmod 777 /tmp/kafka-tp-RouterOnOfflineStateChange-lijieGroup.info
chown zabbix:zabbix /tmp/kafka-tp-RouterOnOfflineStateChange-lijieGroup.info
chmod 777 /home/app/kafka/kafka/bin/kafka-run-class.sh
九、重启zabbix
/etc/init.d/zabbix-agent restart
十、监控上增加3个键值
kafka.offset[RouterOnOfflineStateChange,lijieGroup]
kafka.logsize[RouterOnOfflineStateChange,lijieGroup]
kafka.lag[RouterOnOfflineStateChange,lijieGroup]
原文:https://www.cnblogs.com/lvcisco/p/10957357.html