脚本代码:
#!/bin/bash
#变量初始化
process="php-fpm" #进程名
startCmd="/etc/init.d/php-fpm start" #启动命令
down=0
while true
do
#取得http状态码
code=$(curl -H "Host:www.lsjlt.com" -m 5 -L -s -w %{http_code} http://127.0.0.1 -o /dev/null)
#当状态码返回000或者大于等于500时,计数故障到down变量
if [ $code -eq 000 -o $code -ge 500 ];then
((down++))
else
break
fi
#稍等5s
sleep 5
#判断是否连续检测三次都为故障.
if [ $down -ge 3 ];then
if [ "$(find /tmp/${process}_restart -mmin -3)" == "" ];then
#取得进程名对应的所有pid
pids=$(ps aux | grep ${process} | grep -v "grep" | awk '{print $2}')
#依次对所有pid执行kill命令
for i in $pids;do
kill -9 $i
kill -9 $i
done
#kill完pid后,启动服务
$startCmd
echo "$(date) Return code $code,${process} had been restarted" >> /tmp/${process}_restart
else
echo "$(date) ${process} not yet recovery.As it had been restarted in 2 minutes.so this time ignore." >> /tmp/${process}_not_restart
fi
break
fi
done