1. Proposed title of this feature request

Node Feature Discovery Operator : Allow node-selector and taints configuration for gc components

2. What is the nature and description of the request?

We should be able to configure the pod placement for Node Feature Discovery operator internal components such:

nfd-gc 

3. Why does the customer need this? (List the business requirements here)

We have many customers having limited compute capabilities for running their workloads. They need to schedule internal product component as much as possible on infra nodes.

• They want to limit platform component (except required Daemonset) to run on the compute worker node as much as possible to free their compute capacity
• This specially critic for AI, where GPU nodes are very expensive and they must maximize the GPU nodes usage for AI workload (training and serving) to get a ROI of their GPU nodes pool. Running non AI third-party related component is not an option especially when they have only compute worker nodes having GPU.
• Customers don't want subscription consumption for product components that are not business workload.

4. List any affected packages or components.

Node Feature Discovery Operator