There are a few places which I'd like to revisit to limit how bad we'd be affected by a long timeout set by the user.
One example of this is the stop command. You're able to pass a parameter indicating how long the system should wait before stopping a device. If you run that with a really large time parameter and and shutdown the node controller, then currently in the worst case, assuming everything else runs flawlessly, you'd have to wait about 50 days to see it come back up.
I think it'd be reasonable to limit these parameters for something like 30 minutes or an hour at most for safety reasons.
There are a few places which I'd like to revisit to limit how bad we'd be affected by a long timeout set by the user.
One example of this is the stop command. You're able to pass a parameter indicating how long the system should wait before stopping a device. If you run that with a really large time parameter and and shutdown the node controller, then currently in the worst case, assuming everything else runs flawlessly, you'd have to wait about 50 days to see it come back up.
I think it'd be reasonable to limit these parameters for something like 30 minutes or an hour at most for safety reasons.