Test Volume Replica Zone Soft Anti-Affinity Setting

https://github.com/longhorn/longhorn/issues/5358

Test step - Enable Volume Replica Zone Soft Anti-Affinity Setting

Given EKS Cluster with 3 nodes across 2 AWS zones (zone#1, zone#2)

And Deploy Longhorn v1.5.0

And Disable global replica zone anti-affinity

And Create a volume with 2 replicas, replicaZoneSoftAntiAffinity=enabled and attach it to a node.

When Scale volume replicas to 3

Then New replica should be scheduled

And No error messages in the longhorn manager pod logs.

And Volumes function correctly.

Test step - Disable Volume Replica Zone Soft Anti-Affinity Setting

Given EKS Cluster with 3 nodes across 2 AWS zones (zone#1, zone#2)

And Deploy Longhorn v1.5.0

And Enable global replica zone anti-affinity

And Create a volume with 2 replicas, replicaZoneSoftAntiAffinity=disabled and attach it to a node.

When Scale volume replicas to 3

Then New replica should not be scheduled

When Update the Volume to replicaZoneSoftAntiAffinity=enabled

Then New replica should be scheduled

And No error messages in the longhorn manager pod logs.

And Volumes function correctly.

[Edit]