SGECluster in multiple queues
- Dominant language
- Python
- Stars
- 256
- Forks
- 150
- PR merge metrics
- No merged PRs in 30d
Description
Hi,
I'm opening this issue here, because I'm not having much success with stack overflow (https://stackoverflow.com/questions/59391000/sgecluster-in-multiple-queues). Sorry for the noise.
I'm using `dask.distributed` to launch jobs on a SGE cluster (https://jobqueue.dask.org/en/latest/generated/dask_jobqueue.SGECluster.html#dask_jobqueue.SGECluster) via `dask.bags` and/or `dask.delayed`.
Everything works nicely. However, I may have some `dask.delayed` that should run in a specific queue on my SGE cluster (due to GPU availability for instance (or some other requirement)). How can I work this out on dask?
In other words, how can I define a dask_jobqueue.SGECluster with multiple queues and/or different resource specs? dask_jobqueue.SGECluster allows me to configure one cluster setup only (one queue, one resource spec, etc..).
Thanks
Contributor guide
Assessment
This issue has not been assessed yet.