Discovery process do not continue when Core restarted or aborted

[3keyroman]

Running Dicovery process that is interrupted when the Core is restarted, or because of any reason, will remain in In Progress state and will be never finished.

The Discovery should be checking its state and should completed in with the resulting state even when the process was interrupted.

This can be easily reproduced by:

1. run long Discovery process
2. restart Core
3. the Discovery remains in progress and is never updated

This is observed in 2.12.0.

The solution should be to implement service that is checking the state of Discovery after it is executed. Such independent service should ensure that every Discovery will be eventually processed, whether successfully or not.

Also, it would be great to have a configuration setting in the platform, to set the time after which will be the Discovery automatically marked as failed. Currently, this is hardcoded for 6 hours. It should be configurable. The default value will be taken from setting, but user should be able to override this time for every single Discovery.