MIG support with autoscale to zero is broken

When set directly in the labels before start, the mig configuration process stall in error

An issue was created on the GPU Operator to see if this can be managed upstream. More details of the bug here:

https://github.com/NVIDIA/gpu-operator/issues/2670

Workaround

A workaround was stated here: ryax-engine!45 (closed)

References

• MIG operator doc: https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/latest/gpu-operator-mig.html
• PR on gpu operator: https://github.com/NVIDIA/gpu-operator/pull/2573
• NodeReadinessController: https://github.com/kubernetes-sigs/node-readiness-controller