3.1 years ago by
The usual advice given is to never restart a job. Jobs are never "lost" even if delayed for some reason like a transient cluster issue. These will either execute or fail. There are quotas on how many jobs can be run by any user at any one time. So I wonder if the "24 hrs" includes wait time (grey) while the other jobs processed (yellow to green).
That said, in this specific case, if the job has been executing (yellow) for more than 24 hours that is unexpected. I would go ahead and re-run it and then permanently delete the one executing (this will get it to exit quickly).
If the job takes longer the next pass through, the content is likely a factor. Just let the job run to completion. And consider running FastQC on the original dataset. If you need to double check the correct settings to use with the Groomer, the info here for Fastq Datatype QA should help.
Thanks, Jen, Galaxy team