Using Supervisor to monitor process on CentOS

Prerequisite

Python 2.7 or newer. For CentOS 6.x, it may be necessary to install Python 2.7 from external source.

Install Supervisor with Python 2.7

# First, install the required libraries.
$ sudo yum install -y gcc libffi-devel openssl-devel

# If you're installing Python 2.7 from SCL repo, it is necessary to enable it.
$ . /opt/rh/python27/enable
$ cd /opt/rh/python27/root/usr/bin/

# Install Supervisor and its dependencies.
# For CentOS 6.x
$ sudo LD_LIBRARY_PATH=$LD_LIBRARY_PATH ./pip2.7 install 'requests[security]'
$ sudo LD_LIBRARY_PATH=$LD_LIBRARY_PATH ./pip2.7 install supervisor

# For CentOS 7
$ pip install 'requests[security]'
$ pip install supervisor

Create Supervisor main configuration file

(file: /etc/supervisord.conf)

$ echo_supervisord_conf > /home/sekaimon/supervisord.conf
$ sudo mv /home/sekaimon/supervisord.conf /etc/supervisord.conf
$ sudo sed -i "s/\/tmp\/supervisord.log/\/var\/log\/supervisord\/supervisord.log/g" /etc/supervisord.conf
$ sudo sed -i "s/\;\[include\]/\[include\]/g" /etc/supervisord.conf
$ echo "files = /etc/supervisord.d/*.conf" >> /etc/supervisord.conf

Create the configuration file for the process to monitor

(file: /etc/supervisord.d/selenium.conf)

The following is an example of a configuration for monitoring Selenium grid. Since Selenium grid consists of hub and node processes, it is necessary to monitor those processes as a group. In this example, the Selenium server JAR file is stored in /usr/lib/selenium.

[program:seleniumhub]
command=su <user> -c "java -jar /usr/lib/selenium/selenium-server-standalone.jar -role hub -nodeTimeout 600 > /dev/null 2>&1"
numprocs=1
user=root
stdout_logfile=/var/log/supervisord/selenium.log
stderr_logfile=/var/log/supervisord/selenium.log
redirect_stderr=true
autostart=true
autorestart=true
startretries=0
killasgroup=true
stopwaitsecs=1
priority=998

[program:seleniumnode]
environment=DISPLAY=:99
command=su <user> -c "java -jar /usr/lib/selenium/selenium-server-standalone.jar -role node -hub http://127.0.0.1:4444/grid/register > /dev/null 2>&1"
numprocs=1
user=root
stdout_logfile=/var/log/supervisord/selenium.log
stderr_logfile=/var/log/supervisord/selenium.log
redirect_stderr=true
autostart=true
autorestart=true
startretries=0
killasgroup=true
stopwaitsecs=1
priority=998

[group:selenium]
programs=seleniumhub,seleniumnode
priority=999

Create init script for Supervisor

(file: /etc/init.d/supervisord)

#!/bin/sh
#
# /etc/rc.d/init.d/supervisord
#
# Supervisor is a client/server system that
# allows its users to monitor and control a
# number of processes on UNIX-like operating
# systems.
#
# chkconfig: - 64 36
# description: Supervisor Server
# processname: supervisord

# Source init functions
. /etc/rc.d/init.d/functions

prog="supervisord"

# For CentOS 6.x, it is necessary to enable Python 2.7.
# Otherwise, simply specify the full path to supervisord.
python27_root="/opt/rh/python27"
prog_bin=". ${python27_root}/enable;${python27_root}/root/usr/bin/supervisord"
PIDFILE="/var/run/$prog.pid"

start()
{
       echo -n $"Starting $prog: "
       daemon $prog_bin --pidfile $PIDFILE
       [ -f $PIDFILE ] && success $"$prog startup" || failure $"$prog startup"
       echo
}

stop()
{
       echo -n $"Shutting down $prog: "
       [ -f $PIDFILE ] && killproc $prog || success $"$prog shutdown"
       echo
}

case "$1" in

 start)
   start
 ;;

 stop)
   stop
 ;;

 status)
       status $prog
 ;;

 restart)
   stop
   start
 ;;

 *)
   echo "Usage: $0 {start|stop|restart|status}"
 ;;

esac

Enable the init script and create the necessary directories:

$ sudo chmod +x /etc/init.d/supervisord
$ sudo chkconfig --add supervisord
$ sudo chkconfig supervisord on

$ sudo mkdir -p /etc/supervisord.d
$ sudo mkdir -p /var/log/supervisord

Start supervisord to start monitoring processes

$ sudo service supervisord start

Published

12 April 2016