- Neueste
- Die meisten Stimmen
- Die meisten Kommentare
There's no way to tell AWS Batch to schedule a particular job to a dedicated instance.
However, if you configure your compute environment to spawn EC2 instances that all have the same number of vCPUs, then submitting a job requesting as many vCPUs as are available on your instances will end up running on an instance dedicated to itself (basically, the job will consume all the vCPU of the instance, so batch isn't gonna schedule any additional job on it).
You will loose some flexibility on the instance choice front, but that should fit your requirements.
BR,
Arthur
Ok, answering my own question I managed make this work by creating a compute resource with a specific ec2 instance and always submitting jobs that fully consume the instance's memory and vcpus.
I would be great if this could be achieved pragmatically (like using a flag or something).
Regards.
Relevanter Inhalt
- AWS OFFICIALAktualisiert vor 2 Jahren
- AWS OFFICIALAktualisiert vor einem Jahr