Azure-docs: Can't deploy P100 or V100 in westus2

[ericchansen]

Despite following the YAML templates above, P100 and V100 GPU containers can't be deployed in westus2.

The requested resource is not available in the location 'westus2' at this moment. Please retry with a different resource request or in another location. Resource requested: '4' CPU '28' GB memory 'Linux' OS '1' 'P100' GPU

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: 87729a5b-eda2-985e-678a-45498839f6b8
• Version Independent ID: 8ce5224f-6b82-73e7-15dd-32a146c6ad53
• Content: Deploy GPU-enabled container instance - Azure Container Instances
• Content Source: articles/container-instances/container-instances-gpu.md
• Service: container-instances
• GitHub Login: @dlepow
• Microsoft Alias: danlep

[Karishma-Tiwari-MSFT]

Thanks for the question. We are currently investigating and will update you shortly.

[VikasPullagura-MSFT]

@ericchansen From the error, seems issue with the capacity available in westus2 region when you tried the deployment.
I just tried deploying K80 SKU in westus2 region and it was successful.
As you mentioned P100 & V100 are still not getting deployed due to capacity issues. I tried deploying V100 in centralindia regions and it is successful. You can try to deploy in another supported regions.

[ericchansen]

The documentation should accurately reflect the regions where the resources are available. For example, the docs say that P100 and V100 are available in westus2, which doesn't seem to be the case.

I will try eastus as that is the next closest option for me.

[ericchansen]

It's also not available in "eastus".

[VikasPullagura-MSFT]

@dlepow Can you please check this and add your comments.

[ericchansen]

"westeurope" works.

[VikasPullagura-MSFT]

@dlepow
Can you please check and add your comments on this doc update request as applicable.

[dlepow]

@ericchansen @VikasPullagura-MSFT - Thank you for the feedback. I believe this issue is related to point-in-time availability of GPU resources in the region you tried first, and the error message points to workarounds. The documented resource availability includes a similar note:

"Container groups created within these resource limits are subject to availability within the deployment region. When a region is under heavy load, you may experience a failure when deploying instances. To mitigate such a deployment failure, try deploying instances with lower resource settings, or try your deployment at a later time."

I will update the last sentence to mention trying a different region.

in-progress

[dlepow]

Closing this issue based on doc enhancement. Thanks again for pointing it out.

please-close