一个简单的进程监控脚本

为了保证一个PHP页面在后台运行，写了一个监控脚本，假设程序异常退出，那么可以自动重启。 脚本/usr/local/scripts/receve.sh内容如下：

#!/bin/bash
#funtion:keep  receve.php running
PHP="/usr/local/php5/bin/php"
PROGRAM="receve.php"
#start dameo
$PHP /home/httpd/$PROGRAM >>/usr/local/scripts/phpshell/receve.out &
chpid="$!";
echo "$chpid" > /usr/local/scripts/phpshell/receve.sid
echo "child pid is $chpid"
echo "status is $?"
while [ 1 ]
do
wait $chpid
exitstatus="$?"
echo "child pid=$chpid is gone, $exitstatus" >>/usr/local/scripts/phpshell/receve.php_error.log
echo `date` >> /usr/local/scripts/phpshell/receve.php_error.log
echo "**************************" >>/usr/local/scripts/phpshell/receve.php_error.log
sleep 10
$PHP /home/httpd/$PROGRAM >> /usr/local/scripts/phpshell/receve.out &
chpid="$!";
echo "$chpid" > /usr/local/scripts/phpshell/receve.sid
echo "next child pid is $chpid"
echo "next status is $?"
done

另外也需要保证监控脚本正常运行，在每天凌晨自动重启脚本/usr/local/scripts/restart_r.sh，写入crontab：

#!/bin/bash
echo "" >/tmp/receve_pid.txt
/bin/ps ax | /bin/grep "receve.sh" | /bin/grep -v "grep" | /bin/awk '{print $1}' >/tmp/receve_pid.txt
no1pid=`cat /tmp/receve_pid.txt | sort -r`
no2pid=`cat /usr/local/scripts/phpshell/receve.sid | sort -r`
if [ -z $no1pid ];then
        echo  "receve.sh pid is null, receve.sh is not running."
    else         
        /bin/kill -9 "$no1pid"         
        echo "receve.sh killed" 
fi 
sleep 3
if [ -z $no2pid ];then
        echo  "receve.php pid is null, receve.php is not running."
    else
        /bin/kill -9 "$no2pid"
        echo "receve.php killed"
fi
rm -rf /usr/local/scripts/phpshell/receve.sid 
sleep 3 
/usr/local/scripts/receve.sh > /dev/null &

crontab如下：

#restart receve.sh
10 00 * * * /usr/local/scripts/restart_r.sh

注意：脚本的命名要避免重复，以防把自己给杀了！