Hey this might seem too advanced topic, but it’s so urgent.
I would like to have a 3 machines of my app one of them are serving the actuall app and the other two machine are responsible for ffmpeg encoding commands. as you might know the other two machines won’t run all the time it’ll run only when there is an encoding videos then stops and the first app is running all the time.
The main app would be public with HTTP services on your toml and public IPs. You can edit your toml to make it not auto stop machines or make it stop machines and have at least one running, both cases would satisfy you need for the web server to be always running.
The other app, for ffmpeg, would be similar to the previous app but make it auto stop machines and not have minimum of running machines. Our proxy will stop it when there’s no traffic so you won’t need to worry about it.
Having distinct apps will also make it easier for you on the future to scale vertically and horizontally those differently. I assume you’d want the ffmpeg machines stronger for example.
I have been thinking of this too, if I set the server to auto_stop it’ll stop even if there is an ffmpeg command are executing why is this? and how can I avoid this senario?
Our proxy stops machines with no TCP traffic. If your processing doesn’t require a TCP connection you can disable auto_stop machines but keep auto_start enabled. To stop your machine just exit the process and the machine will be suspended until another request is made.