This PR fixes the issue raised in #86.
At start-up the executions in the database are searched for entries in state "scheduled", "running", or "stopping" which are assumed to be left-overs caused by previously interrupted ndscheduler.
The status of this entries is set to "failed" and the result is set to { "error": "interrupted"}.
This PR fixes the issue raised in #86. At start-up the executions in the database are searched for entries in state "scheduled", "running", or "stopping" which are assumed to be left-overs caused by previously interrupted ndscheduler. The status of this entries is set to "failed" and the result is set to { "error": "interrupted"}.