Timers in persisted machines

[jvalkeal]

This obviously is a very tricky. Lets say that there is a timer which fires and may take machine to some other state. Nothing will happen if machine is not running(as it is hibernated into storage). Should think some ways to do this.

[davidtepel]

We have similar issue - or it is the same one?
We have noticed that the timers are not removed if the machine is removed, i.e creating a new machine instance.
This causes us memroy leaks in the application:

[jvalkeal]

yeah, that's a different issue but I created #321 for it. Tried it and definitely a bug.

[davidtepel]

Thanks for the update

[jvalkeal]

@davidtepel you could actually check if fix(1.2.3 snapshots) for #321 helped with this issue as I made some fixes to stop triggering when machine is stopped.

[davidtepel]

Actually we had to find another solution for the timers becuse of this bug and also, a requirement in which the timeout of the timer should be dynamic. i.e, change the periodic timer timout on the fly without creating a new state machine (in order to change retry mechanism in the same state machine).
So I can't check...