最近项目的看门狗经历了三个版本。
第一个版本:
用ps -ef,如果程序挂了就启动
第二个版本:
程序由于运行时会出现不再监听7901端口,所以不能简单判断机器是不是挂了,而是判断此端口是否有监听
第三个版本:
当7901端口不再监听,就先把原来的killall在启动,每次输出到文件的内容都加日期,要不然根本不知道这事情啥时候发生的
#!/bin/sh set +x source env.sh PRMGRAM=scp_platform FILE_NAME=scp_monitor.log Current_Time=`date +"%Y-%m-%d %H:%M:%S.%N"` echo "[${Current_Time}] monitor start...." echo "[${Current_Time}] monitor start...." >> ${WORK_DIR}/log/${FILE_NAME} port=7905 TCPListeningnum=`netstat -an | grep ":$port " | awk '$1 == "tcp" && $NF == "LISTEN" {print $0}' | wc -l` if [ $TCPListeningnum = 1 ] then { echo "[${Current_Time}] The $port is listening" } else { echo "[${Current_Time}] The port is not listening" } fi while [ 1 ] do Current_Time=`date +"%Y-%m-%d %H:%M:%S.%N"` TCPListeningnum=`netstat -an | grep ":$port " | awk '$1 == "tcp" && $NF == "LISTEN" {print $0}' | wc -l` if [ $TCPListeningnum = 1 ] then { echo "[${Current_Time}] The ${port} is listening" >> ${WORK_DIR}/log/${FILE_NAME} } else { echo "[${Current_Time}] The ${port} is not listening" >> ${WORK_DIR}/log/${FILE_NAME} echo "[${Current_Time}] killall scp_platform now !" >> ${WORK_DIR}/log/${FILE_NAME} kscp echo "[${Current_Time}] check ${PRMGRAM} quit, now restart ${PRMGRAM} ..." >> ${WORK_DIR}/log/${FILE_NAME} scp_platform& } fi sleep 180 done
#bin/bash export ROOT=/root/scp export WORK_DIR=${ROOT} export INCLUDE=${ROOT}/include export OTL=${INCLUDE}/otl_mysql export LD_LIBRARY_PATH=${ROOT}/lib:/usr/local/lib export ACE_ROOT=${INCLUDE} export ODBCINI=/usr/local/etc/odbc.ini export ODBCSYSINI=/usr/local/etc PATH=${PATH}:${ROOT}/bin export PATH odbcinst -j alias wk='cd ${ROOT}' alias bin='cd ${ROOT}/bin' alias cfg='cd ${ROOT}/conf' alias rmlog='rm -rf ${ROOT}/bin/log*.*; rm -rf ${ROOT}/log/*.*' alias lis='netstat -an|grep -i 7905' alias scp='${ROOT}/bin/scp_platform &' alias moni='${ROOT}/bin/monitor.sh &' alias myps='ps -fu root|grep -v grep|grep -i scp' alias mymoni='ps -fu root|grep -v grep|grep -i moni' alias kscp='killall -9 scp_platform' alias kmoni='killall -9 monitor.sh' isql alias mynet='netstat -an | grep 7905' ulimit -c unlimited ulimit -n 65530
原文:http://blog.csdn.net/calmreason/article/details/37921867