InternalOperationError while scaling VMSS

hey hey Mark Carrington & thx for jin me here at Q&A portal,

I assume this looks platform-side. The same scale operation fails through the portal, CLI, and API, while other VMSS changes still work and other scale sets are fine. That points to a stuck or inconsistent backend state on this specific scale set, not a client-side issue. Check the Activity Log and the failed VMSS operation details for a correlation ID and any inner error. InternalOperationError by itself is too generic to fix from the customer side.

I’d avoid repeated scale attempts for now and open an Azure support case with the VMSS resource ID, region, timestamps, correlation IDs, orchestration mode, and whether the failure happens for both manual capacity changes and autoscale.

If the scale set uses Flexible orchestration, mixed VM SKUs, Spot instances, or capacity reservations, mention that too. But the real root cause still needs Microsoft Compute backend logs.

rgds,

Alex

&

If my answer was helpful pls mark it and additional thx if u follow me at Q&A portal 

and at my blog https://ctrlaltdel.blog/

Hello **Mark Carrington

**An InternalOperationError on VM Scale Set (VMSS) scaling typically points to something failing inside Azure or related resources during the update. Since other scale sets work and your VMSS is otherwise configurable (like changing VM SKUs), the most common things to check are locks, unhealthy/partial instance states, transient platform issues, and capacity constraints.

I request you to retry the same scaling action from the portal first, then (if using CLI) retry the operation.

Also, try to verify if any resource locks are existing. Locks can block VMSS operations even when other changes succeed.

• In the Azure portal, go to your VM Scale Set.
• Check Settings > Locks for:
  • the VMSS resource itself.
  • associated resources like disks and network interfaces.
• Look specifically for Read-only or Delete locks.
• Remove any such locks, then retry scaling.

-- If some instances are stuck in problematic states, the scale operation can conflict and fail.

• Go to the VMSS Instances tab. Look for instances in states such as: Failed, Stopped or improperly deallocated states.
• If you find instances that aren’t fully deallocated/healthy, deallocate them and then retry scaling.

az vmss deallocate --resource-group <resource-group-name> --name <vm-scale-set-name>

-- If the scale set is in a bad operational/config state (without it being obvious), redeploying can reinitialize the deployment path.

Portal:

• VMSS page → Support + troubleshooting → Redeploy + Reapply → Redeploy

az vmss update-instances --resource-group <resource-group-name> --name <vm-scale-set-name> --instance-ids '*'

If there’s an ongoing incident impacting compute, scaling may fail while other operations appear normal.

• Check Azure Service Health for compute-related issues in your region.
• If there’s an incident, it’s usually best to retry after it’s resolved.

-- Even if your VMSS is otherwise healthy, scaling up/down might hit capacity constraints depending on region/zone/fault domains and VM size.

If the error details you’re seeing mention capacity, try scaling down to a lower capacity target first, then scale up gradually.

az vmss scale --resource-group <resource-group-name> --name <vm-scale-set-name> --new-capacity <desired-capacity>

(And if you use Availability Zones, ensure your fault domain/platformFaultDomainCount setup is appropriate.)