In current discovery behavior, if two different hosts/network devices coming from two different scanning appliances having same IP address, they get merge or overwrite; if one of the device gets failed on next consecutive discovery, then it can set negative Age Count for a device that is still active/live/accessible.
As being an MSP, there are always changes of two or more devices to share the same IP Address as they belong to different customers. Hence the underlying data coming from each Scanners will be different. If discovery does not consider the company attribute, then it cannot correctly create separate nodes.
Scanner 1 : For customer A with device name abc gets discovered on IP 10.10.10.1
Scanner 2 : For customer B with device name pqr gets discovered on IP 10.10.10.1
at consolidator, these devices will be merge or overwritten
in case machine Customer B's device gets failed for discovery, it will set Age Count to -1.
Solution: BMC Discovery should consider Company in the Host identification process along with the existing unique identification attributes. If there are two hosts getting consolidated for a single IP with different Company name (cdm_company) then these two endpoints should not get merged and be created as separate nodes and synchronized to CMDB.
Discovery should be truly Multi-tenant in discovering and classifying the hosts Tenant wise.