I'm setting up a GCE instance for machine learning training. I first did this:
gcloud compute instances create my_instance --image=pytorch-latest-gpu-v20240128-ubuntu-2004 --image-project=deeplearning-platform-release --accelerator="type=nvidia-tesla-v100,count=1" --custom-extensions --custom-cpu 1 --custom-memory 20GB --boot-disk-size 50GB (some arguments intentionally omitted here)
This command gives a first warning:
WARNING: You have selected a disk size of under [200GB]. This may result in poor I/O performance. For more information, see: https://developers.google.com/compute/docs/disks#performance.
If I modify the above gcloud command to have --boot-disk-size 200GB, then I get a second warning:
- Disk size: '200 GB' is larger than image size: '50 GB'. You might need to resize the root repartition manually if the operating system does not support automatic resizing. See https://cloud.google.com/compute/docs/disks/add-persistent-disk#resize_pd for details.
The link in the first warning is not very helpful. Do I need to be concerned about poor I/O performance warning if I am doing just machine learning model training? I don't want to slow down the model training since the machines are expensive.
