Assign 0 replicas

[LavredisG]

Is there a way to create a PropagationPolicy in a way that assigns 0 replicas to a cluster?
For example if I have 3 replicas and 3 clusters, I would like to set:
cluster1.staticWeight: 3
cluster2.staticWeight: 0
cluster3.staticWeight: 0

but that isn't feasible because a weight of 0 is not valid.

Is there a valid way that makes it work? A workaround would maybe be to delete the cluster affinity of these clusters that are assigned a weight of 0, but I am curious if there is a more straightforward way to do that.

[RainbowMango]

Before answering your question, may I ask the user case under this?

[LavredisG]

Yes ofc. Instead of the the default karmada scheduler I use a custom plugin that after scoring a few possible distributions of replicas, it assigns the best one. So if I want for example to propagate 3 replicas, and I find out that the best scoring distribution is (2,1,0), I would like to set staticWeights to 2:1:0 among the clusters, but that doesn't work due to the last score being 0.

[RainbowMango]

Sorry I didn't make it clear, I want to know the real world use case, like what kind of workloads and why you need to split the replicas at that weight.

[LavredisG]

It's not a "real world" use case, it's for academic project. Essentialy the custom plugin will evaluate a Deployment's possible distributions (input) and after scoring them will choose the best (output) according to our criteria (e.g. power, costs etc). So there happens to be cases where the best distribution has 0 replicas on a cluster.