结合Ansible技术监控Storm集群

1、我的hosts配置
# vim /etc/hosts

192.168.1.100  storm_zk1
192.168.1.101  storm_zk2
192.168.1.102  storm_zk3

192.168.1.103  storm_nimbus

192.168.1.104  storm_supervisor1
192.168.1.105  storm_supervisor2
192.168.1.106  storm_supervisor3
192.168.1.107  storm_supervisor4
192.168.1.108  storm_supervisor5
192.168.1.109  storm_supervisor6

2、我的storm配置
# vim /usr/local/storm/conf/storm.yaml

drpc.servers:
- "storm_supervisor1"
- "storm_supervisor2"
- "storm_supervisor3"

storm.zookeeper.servers:
- "storm_zk1"
- "storm_zk2"
- "storm_zk3"

storm.local.dir: "/data/storm/workdir"

nimbus.host: "storm_nimbus"
nimbus.thrift.port: 6627
nimbus.thrift.max_buffer_size: 1048576
nimbus.childopts: "-Xmx1024m"
nimbus.task.timeout.secs: 30
nimbus.supervisor.timeout.secs: 60
nimbus.monitor.freq.secs: 10
nimbus.cleanup.inbox.freq.secs: 600
nimbus.inbox.jar.expiration.secs: 3600
nimbus.task.launch.secs: 240
nimbus.reassign: true
nimbus.file.copy.expiration.secs: 600
nimbus.topology.validator: "backtype.storm.nimbus.DefaultTopologyValidator"

storm.zookeeper.port: 2181
storm.zookeeper.root: "/data/storm/zkinfo"
storm.cluster.mode: "distributed"
storm.local.mode.zmq: false

ui.port: 8080
ui.childopts: "-Xmx768m"

supervisor.slots.ports:
- 6700
- 6701
- 6702
- 6703
- 6704
- 6705
- 6706
- 6707
- 6708
- 6709

supervisor.childopts: "-Xmx2048m"
supervisor.worker.start.timeout.secs: 240
supervisor.worker.timeout.secs: 30
supervisor.monitor.frequency.secs: 3
supervisor.heartbeat.frequency.secs: 5
supervisor.enable: true

worker.childopts: "-Xmx4096m"
topology.max.spout.pending: 5000

storm.zookeeper.session.timeout: 5000
storm.zookeeper.connection.timeout: 3000
storm.zookeeper.retry.times: 6
storm.zookeeper.retry.interval: 2000
storm.zookeeper.retry.intervalceiling.millis: 30000

storm.thrift.transport: "backtype.storm.security.auth.SimpleTransportPlugin"
storm.messaging.transport: "backtype.storm.messaging.netty.Context"
storm.messaging.netty.server_worker_threads: 50
storm.messaging.netty.client_worker_threads: 50
storm.messaging.netty.buffer_size: 20971520
storm.messaging.netty.max_retries: 100
storm.messaging.netty.max_wait_ms: 1000
storm.messaging.netty.min_wait_ms: 100

3、nimbus节点部署
# vim /data/scripts/monitor_status_for_storm.sh

#!/bin/sh
PATH=/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/bin:/usr/local/sbin

. /etc/profile

## 监控页面地址参数
MON_SRV_IPADDR="192.168.1.103"
MON_SRV_PORT="8080"

## 是否已正确扫描
SCAN_FLAG=0

## 工作基路径
BASE_PATH="/data/scripts"

## 异常 storm Supervisor 主机地址列表
FAIL_SUPERVISOR_LIST="${BASE_PATH}/fail_supervisor.txt"

#---------------------------------------------------------------------------------------------------
## 重启storm的nimbus服务
function restart_storm_nimbus_server()
{
[[ -n `ps aux | grep java | grep storm` ]] && kill -9 `ps aux | grep java | grep storm | awk '{print $2}'`
nohup /usr/local/storm/bin/storm nimbus >/dev/null 2>&1 &
nohup /usr/local/storm/bin/storm ui >/dev/null 2>&1 &

sleep 30
}

#---------------------------------------------------------------------------------------------------
## 1、检查监控页面是否正常【8080端口不通的情况】
for ((i=0; i<3; i++)); do
RETVAL=`/usr/bin/nmap -n -sS -p ${MON_SRV_PORT} ${MON_SRV_IPADDR} | grep open`
[[ -n "${RETVAL}" ]] && SCAN_FLAG=1;break || sleep 10
done

[[ ${SCAN_FLAG} -ne 1 ]] && restart_storm_nimbus_server

#---------------------------------------------------------------------------------------------------
## 2、将监控页面抓取内容与本地hosts内容进行差异比较，以确定是否存在异常的 storm supervisor 服务
curl -s http://${MON_SRV_IPADDR}:${MON_SRV_PORT}/ | sed 's/<td>/<td>\n/g' | awk -F '<' '/^storm_/{print $1}' | awk '!/nimbus/{print}' | sort > ${BASE_PATH}/supervisor_list_from_page.txt

## 如果获取的storm nimbus监控页面数据为空，代表storm nimbus服务存在异常
[[ -z `sed '/^$/d' ${BASE_PATH}/supervisor_list_from_page.txt` ]] && restart_storm_nimbus_server

sort -nr ${BASE_PATH}/supervisor_list_from_page.txt ${BASE_PATH}/supervisor_list.txt | uniq -u > ${BASE_PATH}/supervisor_list_for_failed.txt
[[ -z `sed '/^$/d' ${BASE_PATH}/supervisor_list_for_failed.txt` ]] && rm -f ${BASE_PATH}/supervisor_list_for_failed.txt && exit 0

#---------------------------------------------------------------------------------------------------
## 3、获得异常的 storm supervisor 服务的IP地址列表
echo "[fail_supervisor]" >> ${FAIL_SUPERVISOR_LIST}

for SUPERVISOR_NAMEADDR in `cat ${BASE_PATH}/supervisor_list_for_failed.txt`
do
TEMP_IPADDR=`grep -w ${SUPERVISOR_NAMEADDR} /etc/hosts | grep -v '#' | awk '{print $1}' | tail -1`
echo "${TEMP_IPADDR}" >> ${FAIL_SUPERVISOR_LIST}
IPLIST="${IPLIST} ${TEMP_IPADDR}"
done

#---------------------------------------------------------------------------------------------------
## 4、远程重启 storm supervisor 服务
/usr/local/bin/ansible -i ${FAIL_SUPERVISOR_LIST} fail_supervisor -m shell -a "/data/scripts/restart_storm_service.sh"
rm -f ${FAIL_SUPERVISOR_LIST}

# vim /data/scripts/supervisor_list.txt

storm_supervisor1
storm_supervisor2
storm_supervisor3
storm_supervisor4
storm_supervisor5
storm_supervisor6

# touch /var/run/check_storm.lock
# crontab -e
*/2 * * * * (flock --timeout=0 /var/run/check_storm.lock /data/scripts/monitor_status_for_storm.sh >/dev/null 2>&1)
4、supervisor节点部署
# vim /data/scripts/restart_storm_service.sh

#!/bin/sh
PATH=/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/bin:/usr/local/sbin

. /etc/profile
[[ -n `ps aux | grep java | grep storm` ]] && kill -9 `ps aux | grep java | grep storm | awk '{print $2}'`
nohup /usr/local/storm/bin/storm supervisor >/dev/null 2>&1 &

本文出自 “人生理想在于坚持不懈” 博客，请务必保留此出处http://sofar.blog.51cto.com/353572/1579897