Currently when the SolarJobs application is redeployed (stopped) any executing import jobs can be "abandoned" in the Executing state even though those tasks have been killed. The job state should be handled differently, so they are not left abandoned, which can prevent other jobs with the same group key from starting. For example:

• The task could be cancelled and the job state changed to Complete with a Failed status and error message.
• The task could be cancelled and the job state reset back to Queued so a new SolarJobs instance can pick it up again.