12 months ago by
United States
Hello,
Those executing jobs were already dispatched to a cluster and will eventually end with an error stating that the job was deleted before completing. What you have done (purging) was the correct way to rerun. It may take some time to fully clear out but you didn't have another choice. The clusters are also very busy right now (our admin is monitoring this closely), so some delays are expected for larger job types like Cuffdiff anyway, and a few reruns may not make that much of a time difference overall for your analysis. That said, if you have jobs to run downstream of Cuffdiff, those could also be queued now if getting the work done as quickly as possible is a priority.
Most people have probably started up jobs at some point that have a problem requiring a rerun. The best way to get the new work processed is to leave the rerun jobs queued. They will eventually process and there is no method to speed this up once the deleted/purged jobs are in the execution phase.
FAQs:
- https://galaxyproject.org/main/
- https://galaxyproject.org/support/how-jobs-execute/
Thanks! Jen, Galaxy team
