Status 'waiting'

Status 'waiting'  

  By: ZoeRen on March 4, 2024, 5:04 p.m.

Hi,

I run my validation.py with an Ubuntu-Python-TensorFlow image, but always receive a status of 'waiting' from Jobman. Can anyone help me?

Re: Status 'waiting'  

  By: alvaroparicio on March 5, 2024, 2:34 p.m.

Hi, it's the normal operation of the platform. There are 4 types of resources configuration (flavors) and the resources for each type are different, so it can be seen as 4 queues. If you submit a job with the "no-gpu" flavor (the default) and the job stay in waiting status, because there are other jobs of other users already in the queue, you can wait, or you can cancel the job and select another flavor (if your job fits to it) to go to other queue that can be empty depending on the demand of the people in that moment (the command "jobman queue" gives you an idea). Also take into account, some queues move faster, like the "small-gpu" due to the limit of job execution time of 12h and the purpose (usually it is used to test and debugging jobs), and others like "large-gpu" moves slower again due the same reasons (the oposite case: 24h of limit, and large jobs usually sent).

Hope we could help!