On the parent NS, if you change the resource status from Active to something else (like Retired, Out of order, etc), when replication occurs it is overwritten by whatever status that resource has in the Child NS. This is part of the original design where the Child NS is the source of truth regarding resource status values. However, if you are using Asset/CMDB on the parent and you need to set a different resource status, this is ignored.
Also, if Basic Inventory is received for any computer resource that has a custom non-active status, that status is overwritten and changed to Active.
Working as designed. The purpose of a hierarchy is for objects to be replicated between parent and child servers. This would include the properties, status, and state of the objects. However, there is a workaround to prevent the core services from switching the status of a computer to the status it had received via replication or basic inventory.
Steps to disable overwriting the status when resources are replicated from the Child NS to Parent NS:
Steps to re-enable overwriting the status when resources are replicated from the Child NS to Parent NS:
See also KB 152868 "'Retired' status is being changed automatically to 'Active' when basic inventory reports in."