Graceful shutdown for Orchestrators
Is your feature request related to a problem? Please describe. Orchestrators can't currently shut down without breaking active sessions
Describe the solution you'd like CLI option to tell an O to stop accepting new sessions and to shut itself down cleanly when current sessions have finished
Describe alternatives you've considered Ideally this will happen automatically as part of a broader node upgrade automation, but this is a good first iteration.
I'd like the ability to gracefully shut down a T on a split O/T setup too.
Today I wanted to update the OS on 2 transcoders connected to my LAX O both of which have had sessions all day. If I could have gracefully shut one down, updated the OS, then reconnected it, and done the same with the other afterward, I wouldn't be sitting here thinking should I break the sessions on one the Ts or wait until there's a natural break in work.
I wouldn't call this a high priority issue but it'd be a nice to have.