Here is what I am trying to accomplish from design perspective for a multi-site setup.
Logically, I have one tenant, one VRF and one App Profile in DC-1. Say, there are about 10 EPGs (1 EPG=1 VLAN=1 BD) in place.
#1, I want to perform cold migration of workloads under EPG-1 to EPG-3 to my DC-2 for the "one or nothing" disaster-recovery scenario (workloads need to maintain the same IP addresses). So I would stretch the EPG-1 to EPG-3 with associated BDs between two DCs without flooding L2 BUM frames inter-site.
#2, I also want to setup DC-2 specific/local AP/EPG/BD which can only reach the live workloads under the existing EPGs in DC-1 via L3 Routing.
Referencing to Multi-Site whitepaper below, it seems like #2 above is fully supported but
a. In order to support my #1 above, my tenant, VRF, EPGs/BDs are all required to be stretched across. So I can not ONLY stretch 3 of my 10 EPGs/BDs while keep the remaining 7 EPGs/BDs locally to DC-1?
b. I did not get a clear understanding on the supportiveness of both #1 and #2 use cases in the same setup?
https://www.cisco.com/c/en/us/solutions/collateral/data-center-virtualization/application-centric-infrastructure/white-paper-c11-739609.html#CiscoACIMultiSiteperbridgedomainbehavior
Suggestions?
