Simple implementation of linux process monitoring and automatic restart

Purpose:

The server program under linux will drop dump for various reasons, which will affect user use. Here is a simple process monitoring and restart function.

Implementation principle:

The script is invoked by the timer task crontab, which checks for the existence of the process using ps, and restarts and writes to the log if it does not exist.

crontab Modification

# crontab -e
*/5 * * * * /mnt/bindmonitor.sh

Implementation of/mnt/bindmonitor.sh

#! /bin/sh

host_dir=`echo ~`          #  Current User Root Directory 
proc_name="/home/wkubuntu/named/sbin/named"        #  Process name 
file_name="/mnt/bindmonitor.log"       #  log file 
pid=0

proc_num()            #  Calculate the number of processes 
{
 num=`ps -ef | grep $proc_name | grep -v grep | wc -l`
 return $num
}

proc_id()            #  Process Number 
{
 pid=`ps -ef | grep $proc_name | grep -v grep | awk '{print $2}'`
}

proc_num
number=$?
if [ $number -eq 0 ]         #  Determine whether a process exists 
then 
 /home/wkubuntu/named/sbin/named -c /home/wkubuntu/named/etc/named.conf -n 1 &
              #  The command to restart the process, please modify it accordingly 
 proc_id           #  Get the new process number 
 echo ${pid}, `date` >> $file_name  #  Record new process number and restart time 
fi

Delete Process Tests

a. #killall -15 named

After b. 5 minutes, cat/mnt/bindmonitor.log checks to see if there is an up-to-date record. The process number corresponds to the # ps-ef | grep named process number.

summary