Unexpectedly long queuing for task

#1

Hi,

I launched a simple task from a collab jupyter notebook. This type of task usually starts running withing five minutes, but this time it is still in the queue over 45 minutes after being submitted. Is this expected? If not, could somebody have a look? The job ID is a9f8e44e-2be9-11e6-9228-0050569773f6 and it was launched with the following parameters:

{'total_physical_memory': 4048, 'requested_queue': 'cscs_viz_webruns'}

A corollary question is: is there a means to check the queues from a collab?

Cheers,

Juan

#2

The task has been in the queue for over 18 hours now. This would seem excessive. I’d like to know

  • If this is really an issue or is something that can be expected under normal circumstances
  • If it is an issue, what kind of timescale should we expect to have it addressed

I am asking because this is part of our M30 work flow and will be part of the so-called EU “review” supposedly taking place this month. So if we can expect such delays, either of normal running of for addressing potential issues, we need to document this and make it clear to the reviewers that the process may take days rather than minutes.