They added a domain, added the APIC controller, added VNA to Spectrum and PM, all looks good. What happens when the APIC goes down? How can he guarantee maximum monitoring capabilities? Is this something that has to be guaranteed on the APIC side with failover setup on a IP address or do we have to add multiple APICs? Any architectural guidance on this?
Release : 20.2
Component : MITS SPECTRUM INTEGRATION
We learn about all APICs in the cluster when we connect to it. If a single APIC goes down, we will try to contact one of the others. So when a customer configures an ACI Engine with the IP of one of the APICs ( say 126.96.36.199 ) we will learn about the other APICs ( say 188.8.131.52 and 184.108.40.206 ) and if 220.127.116.11 goes down, we will try the other two IPs. We learn about those from the APIC's APIs.