Cisco UCS Management IP Address assigned twice?

Question

I have a blade chassis with 8 servers, and a pool of management addresses assigned to a service profile template.  All 8 servers are built from this template.  There are 8 addresses in the pool.
I am finding that 4 servers in I have exhausted all 8 addresses from the management pool.  Upon further investigation, it looks like all servers have taken 2 addresses from this pool.  One address is the address assigned to org-root/service-profile-name/ipv4-pooled-addr.  Another address (same server) is assigned to sys/chassis-number/blade-number/mgmt/ipv4-pooled-addr.
What am I doing wrong?

Answer 1

It is my understanding that this is the correct / default behaviour. When you created a mgmt pool, physical hardware automatically picks up an address after discovery. When you create a template, you do not necessarily need to assign the pool to the template as when you accociate a profile with hardware then you can access the KVM to the profile via the address assigned to the hardware. The upside of assigning a pool to the template / profile is the fact that that address will always live with the profile (server) as opposed to an address on the hardware which could have different profiles assigned at different times.  So, if you assign to the template / profile you will use 2 x the addresses from the pool.