Supervisor: it keeps your app running
Check status
service supervisor status
Tail a service's err
(find the file by looking in folder: /etc/supervisor/conf.d where file pdf.nimbletext.com.conf might mention error log /var/log/pdf.nimbletext.err.log)
sudo tail -n 20 -f /var/log/pdf.nimbletext.err.log
Restart supervisor
service supervisor stop
service supervisor start
supervisor is a nice thing. It supervises your app to keep it running.
sudo apt-get update
sudo apt-get install supervisor
Config goes in...
cd /etc/supervisor/conf.d/
To create a new service for supervisor to supervise, add a file, e.g. helloworld.conf in the directory mentioned above, with content such as
$ cat helloworld.conf
[program:helloworld]
command=/usr/bin/dotnet /home/user1/publish/MvcMovie.dll
#var/aspnetcore/HelloMVC/HelloMVC.dll
#directory=/var/aspnetcore/HelloMVC/
directory=/home/user1/publish/
autostart=true
autorestart=true
stderr_logfile=/var/log/helloworld.err.log
stdout_logfile=/var/log/helloworld.out.log
environment=ASPNETCORE_ENVIRONMENT=Production
user=www-data
stopsignal=INT
View supervisor logs
Here's how to view the last 20 lines of the supervisor log, using tail
sudo tail -n 20 -f /var/log/supervisor/supervisord.log
...in which you might see it struggle to keep your app running...
Or the last 20 lines of your app's error output...
sudo tail -n 20 -f /var/log/helloworld.err.log
Or the last 20 lines of your app's standard (non error) output...
sudo tail -n 20 -f /var/log/helloworld.out.log
Restart supervisor - troubleshooting
To restart supervisor after you've given it some new configuration:
service supervisor stop
service supervisor start
I noticed that supervisor did not successfully restart when I rebooted my linux box.
I restarted it with service supervisor start (which required me to type my password)
First though, I checked the status with:
service supervisor status
and saw:
user1@machine1:~$ service supervisor status
? supervisor.service - Supervisor process control system for UNIX
Loaded: loaded (/lib/systemd/system/supervisor.service; disabled; vendor preset: enabled)
Active: inactive (dead)
Docs: http://supervisord.org
Starting it went like this...
user1@machine1:~$ service supervisor start
==== AUTHENTICATING FOR org.freedesktop.systemd1.manage-units ===
Authentication is required to start 'supervisor.service'.
Authenticating as: user1,,, (user1)
Password:
==== AUTHENTICATION COMPLETE ===
Check status of supervisor
And now I get this result from status --
user1@machine1:~$ service supervisor status
? supervisor.service - Supervisor process control system for UNIX
Loaded: loaded (/lib/systemd/system/supervisor.service; disabled; vendor preset: enabled)
Active: active (running) since Thu 2017-04-06 07:19:51 EDT; 4min 19s ago
Docs: http://supervisord.org
Main PID: 2795 (supervisord)
Tasks: 41
Memory: 275.4M
CPU: 24.093s
CGroup: /system.slice/supervisor.service
+-2795 /usr/bin/python /usr/bin/supervisord -n -c /etc/supervisor/supervisord.conf
+-2800 /usr/bin/dotnet /opt/webapps/webapp1/app/app1.dll
+-2801 /usr/bin/dotnet /opt/webapps/webapp2/app/app2.dll
+-2802 /usr/bin/dotnet /opt/webapps/webapp3/app/app3.dll
Another way to inspect the status of supervisor:
systemctl status supervisor.service
Exactly same output as service supervisor status...e.g.:
? supervisor.service - Supervisor process control system for UNIX
Loaded: loaded (/lib/systemd/system/supervisor.service; disabled; vendor preset: enabled)
Active: active (running) since Mon 2017-09-11 02:00:44 UTC; 1min 27s ago
Docs: http://supervisord.org
Main PID: 6654 (supervisord)
Tasks: 41
Memory: 207.6M
CPU: 14.514s
CGroup: /system.slice/supervisor.service
+-6654 /usr/bin/python /usr/bin/supervisord -n -c /etc/supervisor/supervisord.conf
+-6659 /usr/bin/dotnet /opt/webapps/yourfirstproduct.com/app/YourFirstProduct.dll
+-6660 /usr/bin/dotnet /opt/webapps/codesmackdown.com/codesmackdown.com.dll
+-6661 /usr/bin/dotnet /opt/webapps/joy.nimbletext.com/app/Joy.dll
Make supervisor restart automatically
To restart supervisor on reboot, add something like this to your crontab:
@reboot /my/path/to/supervisord -c /my/path/to/supervisord.conf 2>&1
i.e.
@reboot /usr/bin/supervisord -n -c /etc/supervisor/supervisord.conf 2>&1
Add that to your cron jobs using crontab (described in cron).
Useful .ok file for /etc/supervisor/conf.d folder
I use ok.sh on all my linux/mac systems (and ok-ps on all my windows systems)
ls *.conf -al # view config of each app
cat *.conf | grep 'out\.log\|program'
sudo tail -n 100 -f /var/log/supervisor/supervisord.log # check sup's log
sudo tail -n 150 -f /var/log/myapp1.out.log
sudo tail -n 150 -f /var/log/myapp2.out.log
service supervisor status
cd /etc/nginx/; d # nginx
(Those lines that say for example: sudo tail -n 150 -f /var/log/myapp1.out.log ...i have something other than myapp1 -- i have the name of the log files output by my web apps. The same files I can find the name of with cat *.conf | grep 'out\.log\|program')
The last line of the .ok file takes me to the nginx config folder, where I have another .ok file.