Installing and Configuring Supervisor on CentOS 8

Traducciones al Español
Estamos traduciendo nuestros guías y tutoriales al Español. Es posible que usted esté viendo una traducción generada automáticamente. Estamos trabajando con traductores profesionales para verificar las traducciones de nuestro sitio web. Este proyecto es un trabajo en curso.
Create a Linode account to try this guide with a $ credit.
This credit will be applied to any valid services used during your first  days.

Supervisor is a process control system based on the client/server model. It can be used to simplify process management by providing a centralized location for process control. It’s most often deployed to control services that don’t have initialization, auto-start, or management scripts. Remote process control is also supported via Remote Procedure Calls, or RPC.

As an example, if you have written a custom Node.js web application, Supervisor could be used to ensure that it starts on boot. As well, Supervisor could restart the application if it quits unexpectedly.

Note
This guide uses a Python program called app.py as an example for process control. Supervisor can control Python applications, Node.js applications, or programs written in other languages or runtimes.

In this Guide

This guide shows how to:

Before You Begin

  1. If you have not already done so, create a Linode account and CentOS 8 Compute Instance. See our Getting Started with Linode and Creating a Compute Instance guides.

  2. Follow our Setting Up and Securing a Compute Instance guide to update your system. You may also wish to set the timezone, configure your hostname, create a limited user account, and harden SSH access.

    Note
    Commands that require elevated privileges are prefixed with sudo. If you’re not familiar with the sudo command, see the Users and Groups guide.
  3. Install Python 3.6 or newer.

Set Up the Example App

Supervisor is usually configured to run important components of your project’s software. The example in this guide is simplified so that it just illustrates Supervisor’s functions:

  1. Log into your CentOS 8 server.

  2. Create the directory that the application is stored in:

     sudo mkdir /opt/myapp/
    
  3. Create the app.py program file inside this directory with the text editor of your choice. You need to use sudo with your text editor to create the file. For example, to create it with the nano text editor, run:

     sudo nano /opt/myapp/app.py
    
  4. Paste the content of this snippet into the file:

    File: /opt/myapp/app.py
     1
     2
     3
     4
     5
     6
     7
     8
     9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    
    #!/usr/bin/python3
    
    import sys
    import random
    import time
    from datetime import datetime
    
    failure_chance = .01
    sleep_interval = .1
    sleep_counter = 0
    
    # Execute loop until random() generates a number less than the failure chance
    # that was specified. Also, don't exit the loop in the first second, even
    # if a random number is generated under the failure chance during the first second
    while random.random() > failure_chance or sleep_counter * sleep_interval < 1:
        time.sleep(sleep_interval)
        sleep_counter += 1
    
        # Print the current date/time and the number of seconds that the program
        # has been running for to the system's standard output
        print(f"{datetime.now().strftime('%Y-%m-%d %H:%M:%S')}: Alive for {round(sleep_counter * sleep_interval,1)} seconds", file = sys.stdout)
        sys.stdout.flush()
    
    print("---", file = sys.stdout)
    
    # Print the current date/time and the number of seconds that the program
    # ran the loop for to the system's standard error output
    print(f"{datetime.now().strftime('%Y-%m-%d %H:%M:%S')}: Died after {round(sleep_counter * sleep_interval,1)} seconds", file = sys.stderr)

    This program will run for a random amount of time before quitting. Each cycle of the program’s main loop will have a chance to exit equal to the failure_chance on line 8. Each time the program’s main loop runs, it will output how long the loop has been running to the system’s standard output. At the end, it will output how long the program ran to the system’s standard error output.

  5. Set your myappuser user as the owner of the program directory:

     sudo chown -R myappuser:myappuser /opt/myapp
    
  6. Make the new program executable:

     sudo chmod u+x /opt/myapp/app.py
    

Install and Configure Supervisor

Supervisor is in the EPEL package repository for CentOS. Install this package repository, then install Supervisor:

sudo yum install epel-release
sudo yum install supervisor

(Optional) Configure Supervisord.service

If you intend to use Supervisor to manage processes that require the network to be online, edit the /usr/lib/systemd/system/supervisord.service file. Add network-online.target to the After parameter of the [Unit] section. For example:

File: /usr/lib/systemd/system/supervisord.service
1
2
3
4
5
[Unit]
Description=Process Monitoring and Control Daemon
After=rc-local.service nss-user-lookup.target network-online.target

...

(Optional) Set the Location of Process Configuration Files

In this guide, individual processes have their own .ini configuration files placed in /etc/supervisord.d/. This location is specified in /etc/supervisord.conf by default:

File: /etc/supervisord.conf
1
2
[include]
files = supervisord.d/*.ini

If you prefer to store configuration files elsewhere, edit the [include] section of /etc/supervisord.conf to point to the directory of your choice.

Add Process Configuration Files

Process configurations can be added directly in /etc/supervisord.conf. However, configuration management can be simplified by placing individual configuration files in /etc/supervisord.d/.

  1. Use sudo to create a myapp.ini configuration file in /etc/supervisord.d/. Paste this snippet into the new file:

    File: /etc/supervisord.d/myapp.ini
     1
     2
     3
     4
     5
     6
     7
     8
     9
    10
    11
    12
    
    [group:myappgroup]
    programs=myapp
    
    [program:myapp]
    directory=/opt/myapp                     ; Location of application
    command=python3 app.py                   ; The command to execute
    autostart=true                           ; Start this application when supervisord starts
    autorestart=true                         ; Restart this application if it crashes
    stderr_logfile=/var/log/myapp/app.err.log  ; Make sure this directory exists
    stdout_logfile=/var/log/myapp/app.log      ; Make sure this directory exists
    stopsignal=INT                           ; Signal sent to the application when halting
    user=myappuser                           ; setuid to this UNIX account to run the program

    The configuration file defines a group. Groups help you manage several processes as a single unit from Supervisor.

    It then defines a program, with several settings that control when and how the program is run.

  2. The example program configuration provides logging of stderr and stdout to locations specified in stderr_logfile and stdout_logfile. The directory specified for these logs (/var/log/myapp/) needs to be created before starting/restarting supervisord. Create this directory:

     sudo mkdir /var/log/myapp
    
    Note

    If you do not create this directory, the following error appears when you try to load the new Supervisor configuration:

    ERROR: CANT_REREAD: The directory named as part of the path /var/log/myapp/app.log does not exist in section 'program:myapp' (file: '/etc/supervisord.d/myapp.ini')

Run Supervisor

  1. Enable and start the supervisord service:

    sudo systemctl enable supervisord
    sudo systemctl start supervisord
    
  2. Once supervisord has been started you can access it via the command line with the supervisorctl command. Run the status action for supervisorctl to check the uptime of your program:

    sudo supervisorctl status
    
    myappgroup:myapp                 RUNNING   pid 16407, uptime 0:00:09
  3. You can invoke the supervisorctl command on its own to bring up a command prompt:

    sudo supervisorctl
    
    myappgroup:myapp                 RUNNING   pid 18833, uptime 0:00:10
    supervisor> status

    Once this prompt appears, you can type the name of action to perform it:

    status
    
    myappgroup:myapp                 STARTING

    Enter quit to leave the prompt.

  4. You can also see that logs are being written to the log files specified earlier:

    tail /var/log/myapp/app.log
    
    2021-02-22 16:11:46: Alive for 4.3 seconds
    2021-02-22 16:11:46: Alive for 4.4 seconds
    2021-02-22 16:11:46: Alive for 4.5 seconds
    2021-02-22 16:11:46: Alive for 4.6 seconds
    ---
    2021-02-22 16:11:47: Alive for 0.1 seconds
    2021-02-22 16:11:47: Alive for 0.2 seconds
    2021-02-22 16:11:47: Alive for 0.3 seconds
    2021-02-22 16:11:47: Alive for 0.4 seconds
    2021-02-22 16:11:47: Alive for 0.5 seconds
    tail /var/log/myapp/app.err.log
    
    2021-02-22 16:12:22: Died after 13.1 seconds
    2021-02-22 16:12:30: Died after 6.2 seconds
    2021-02-22 16:13:16: Died after 45.0 seconds
    2021-02-22 16:13:23: Died after 6.4 seconds
    2021-02-22 16:13:26: Died after 2.0 seconds
    2021-02-22 16:14:01: Died after 32.4 seconds
    2021-02-22 16:14:05: Died after 3.0 seconds
    2021-02-22 16:14:08: Died after 1.7 seconds
    2021-02-22 16:14:15: Died after 5.6 seconds
    2021-02-22 16:14:26: Died after 10.7 seconds
    Note
    Optionally, you can install and use the logrotate tool to manage the log files created by Supervisor.

Other supervisorctl Actions

Here’s the syntax for a few other basic commands:

  1. Stop your program with the stop action. Specify the group it belongs to and the program name:

     sudo supervisorctl stop myappgroup:myapp
    
    myappgroup:myapp: stopped
  2. Reference all of the programs in a group with the groupname:* syntax. Run this start action to start all the programs in the myappgroup group (which is just the single example program in this guide):

     sudo supervisorctl start myappgroup:*
    
    myappgroup:myapp: started
  3. Restart programs with the restart action:

     sudo supervisorctl restart myappgroup:myapp
    
    myappgroup:myapp: stopped
    myappgroup:myapp: started
  4. If you have made a configuration change to a Supervisor process, you can run the reread command to reload the changes:

     sudo supervisorctl reread
    
  5. The reread command does not restart programs. Run the update command to both reload any configuration changes and restart programs:

     sudo supervisorctl update
    

A full list of actions and other supervisorctl documentation can be found at Running Supervisor — Supervisor 4.2.1 documentation.

Enabling HTTP Access (Optional)

You may want to add HTTP access to supervisord, either to enable the web interface or to allow remote RPC calls.

Important
Enabling HTTP access exposes supervisord to the internet at large. If you choose to enable HTTP access, make sure to configure firewall rules that limit access to trusted IPs. As well, configure a user name and a long, complex, and unique password for service access.

  1. To enable HTTP access, uncomment the [inet_http_server] in /etc/supervisord.conf. Update the port, username, and password settings:

    File: /etc/supervisord.conf
    1
    2
    3
    4
    
    [inet_http_server]
    port=*:9001              ; IP address and port to bind to. Use *:9001 to listen on all interfaces.
    username=super                 ; Service user name
    password=A!VeryS3cuReP@5sw0rd  ; Service password, make it a good one.
  2. Enable a firewall rule to allow access for your remote IP or a trusted network. Replace the value for address with your IP address or network range:

    sudo firewall-cmd --permanent --zone=public --add-rich-rule='
      rule family="ipv4"
      source address="4.3.2.1/32"
      port protocol="tcp" port="9001" accept'
    sudo firewall-cmd --reload
    
  3. Restart supervisord

     sudo systemctl restart supervisord
    
  4. You should now be able to visit http://<yourIPOrDomain>:9001 and log in with the configured user name and password. A listing of your programs appears. This interface allows you to view your programs’ uptime, restart and stop your programs, and view the programs’ logs:

More Information

You may wish to consult the following resources for additional information on this topic. While these are provided in the hope that they will be useful, please note that we cannot vouch for the accuracy or timeliness of externally hosted materials.

This page was originally published on


Your Feedback Is Important

Let us know if this guide was helpful to you.


Join the conversation.
Read other comments or post your own below. Comments must be respectful, constructive, and relevant to the topic of the guide. Do not post external links or advertisements. Before posting, consider if your comment would be better addressed by contacting our Support team or asking on our Community Site.
The Disqus commenting system for Linode Docs requires the acceptance of Functional Cookies, which allow us to analyze site usage so we can measure and improve performance. To view and create comments for this article, please update your Cookie Preferences on this website and refresh this web page. Please note: You must have JavaScript enabled in your browser.