I have a n1-standard-8 (8 vCPUs, 30 GB memory) instance on Google Cloud with a 600GB standard persistent disk attached to it. It works normally, until I try to attach a GPU (NVIDIA Tesla K80) to it and restart it. Then I get the following message:
Starting VM instance "instance_name" failed. Error: The zone 'projects_name/zones/us-east1-d' does not have enough resources available to fulfill the request. Try a different zone, or try again later.
I've been getting this error for the past two weeks. For a while I assumed that this was an issue with GPUs being occupied by other users on the same zone. So I moved the instance (along attached disk) to a different zone using gcloud compute instances move with no difference in the outcome. Cloning the instance on other zone/regions did not change the error either.
But I'm guessing that this is not really a resource issue, since I am able to create a new instance with the same specs (CPU/GPU/Memory) on the same zone and run it without any problem.
I was able to attach GPUs to this instance about one or two months ago. But a lot of resizing, and coding has been done since then. So I'm not sure if I have changed some important settings that could be relevant. Does anyone know why this is happening?
What is the recommended solution?
-Thanks
