This page describes problems that can come up when creating Deep Learning VM Images instances, and tells you how to address the problems.
- Quota 'NVIDIA_K80_GPUS' exceeded. Limit: 0.0 in region
Problem: You do not have enough quota.
Solution: You must have GPU quota before you can create instances with GPUs. Check the quotas page to ensure that you have enough GPUs available in your project. If GPUs are not listed on the quotas page or you require additional GPU quota, request a quota increase. If your project has an established billing history, it will receive quota automatically after you submit the request. Free Trial accounts do not receive GPU quota by default.
Keep in mind that preemptible GPUs and normal GPUs require separate quota requests. You can't use preemptible GPU quota for normal GPUs. Also, quota is per region, so be sure that you are creating the instance in the region where you have quota.
Resource not found
- The resource 'projects/deeplearning-platform/zones/europe-west4-c/acceleratorTypes/nvidia-tesla-k80'
was not found
Problem: You are trying to create an instance with one or more GPUs in a region where GPUs are not available (for example, an instance with a K80 GPU in europe-west4-c).
Solution: To determine which region has the required GPU, see GPUs on Compute Engine.
Symptom: I can't create preemptible instance from the UI, even though I have quota.
Solution: At this time, a preemptible instance can't be created from
Google Cloud Marketplace. You must use the CLI. Be sure to add
when setting up your new instance.
Unable to use SSH port forwarding to connect to JupyterLab
Symptom: When using SSH port forwarding to connect to JupyterLab, you are unable to connect to the instance.
Problem: You are trying to connect to the wrong TCP socket.
On some Linux clients, the localhost resolves to the IPv6 loopback address (
::1). Check this by using
ping -c 1 localhost. If this command returns the IPv6 address, use
-L 8080:127.0.0.1:8080(rather than
-L 8080:localhost:8080) in the
gcloud compute sshcommand.
Ensure that you connect to
https://localhost:8080) on your local client.