VM allocation Failing - unable to start machine

Question

We are unable to start existing machine getting error of "Failed to start virtual machine 'GPU-machine'. Error: Allocation failed. We do not have sufficient capacity for the requested VM size in this zone. Read more about improving likelihood of allocation success at http://aka.ms/allocation-guidance".

Machine size is Standard_NC40ads_H100_v5 in US west 2.

Answer 1

Hello @Neelesh Kankane ,

This is a known and expected behavior in Azure.

It indicates that the required physical hardware for your selected VM size is temporarily unavailable in west US. Azure VM capacity depends on the availability of physical hardware within each region and availability zone. When a specific zone runs out of capacity for a particular VM SKU, new deployments or allocations for that size cannot be completed.

Not to impacting your business operations since the VM is required for critical workloads. To avoid encountering the same issue in the future, it is recommended to select multiple availability zones during deployment, which helps improve resiliency and increases the likelihood of capacity availability.

Please note that this is a temporary capacity constraint and is not caused by any misconfiguration in your environment. Sometimes redeploy can resolve the allocation issues.

Answer 2

Hello @Neelesh Kankane,

This situation can occur due to high demand for certain VM sizes in specific regions or zones. To resolve this issue, you can try the following workarounds:

1. Retry the allocation: Sometimes, the issue may be temporary, and retrying after a short period can help.
2. Resize the VM: Consider changing to a different VM size that may have more availability in the region or zone. For example, you could try alternative sizes like Standard_A2_v2 or Standard_D2a_v4.
3. Change the region or zone: If the current zone is experiencing high demand, deploying the VM in a different zone or region might provide more capacity.

---

If this answers your query, do click Accept Answer and Up-Vote for the same. And, if you have any further query do let us know.

Answer 3

Hello Neelesh

Thank you for contacting us about the VM issue.

We understand how impactful this issue can be, especially when working with GPU‑based workloads. The error you are encountering indicates a temporary Azure platform capacity limitation for the selected VM size in the chosen region or availability zone. This is a known behavior for high‑demand GPU SKUs and is not related to your VM configuration or subscription limits.

As outlined in Microsoft’s official guidance, we recommend retrying the operation after some time, attempting a different availability zone or region, or considering an alternative VM size if your workload allows. For workloads that require consistent availability, planning ahead with On‑Demand Capacity Reservations can help improve allocation reliability.

https://learn.microsoft.com/en-us/troubleshoot/azure/virtual-machines/windows/allocation-failure

The above document explicitly recommends the following supported actions:

1. Retry the allocation
   • Capacity may become available as resources are freed in the cluster or zone.
2. Change the Availability Zone or remove zonal constraint
   • Expands allocation options beyond a single zone.
3. Resize to an alternative VM size
   • If workload permits, selecting a different SKU may succeed.
4. Deploy or migrate the VM to another region
   • Recommended when the current region is experiencing high demand.
5. Use On‑Demand Capacity Reservations
   • Helps ensure compute availability for future start or redeployment operations.

Workaround suggested : As capacity is dynamic and Capacity is currently constraint for the SKU and try deploying NDsr_H100_v5-series in the same region.

If the issue persists despite following the recommended guidance, please feel free to comment in the Comment section, and we will be happy to review your scenario further and assist you promptly.

Thanks,
Manish.