5 days certainly doesn’t sound right…although I have heard of cases where when the job starts, it immediately fails but doesn’t notify the job-scheduler of the failure, and makes it look like the job is hung/infinitely running.
To restart/reset all jobs, you can restart the WebAPI service. That will terminate any running job.
For your specific environment, we need a bit more information:
What version of WebAPI are you running
What database is your CDM on (oracle, redshift, sql server)
Do you have access to a console that you can execute Queries? I can show you where to go to run the pathway analysis queries directly so you can narrow down where your specific problem is.
If you have access to the WebAPI logs, you may find information in there about a failure/problem, but if it was already running for 5 days, it could be buried. Better to just restart the service, generate again, and then look at the logs.