I might not use the correct terminology, but here below are what I am confused about by reading Cisco design doc for the solution...
According to Cisco SD-Access Solution Design Guide, it states "Typically, fabric WLCs connect to a shared services network though a distribution block or data center network that is connected outside the fabric and fabric border, and the WLC management IP address exists in the global routing table." But within the SD-Access Wireless Design and Deployment Guide, Cisco DNA Center 2.1.1, it also states "The fabric-enabled WLC can only manage a single fabric site." and "The WLC and APs need to be within 20 ms of latency. Usually, this means being on the same physical site."
So which guideline should I follow? Is 20-ms really a hard requirement? If not, should I put WLC in shared data center OR onsite? If in shared DC, I would have to put 20 of them there if I have 20 fabric site?
For fabric enable wireless, it requires separate wlc per fabric site. If you are planning to have one large fabric site then you can use single wlc but if you are planning 20 different fabric site then it requires that many wlc. You will still need to meet latency requirement from each site to have better user experience or you can put wlc in shared services upstream to border or you can use fiab with wireless on Cat9k switch depending upon scale of the site.
So what you are saying is both those statements from the two documents are correct as "a fabric WLC can be installed onto data center network but it needs to have less than 20-ms lactency to the Fabric APs onsite and it can only manage one site"?
If this is the case, say I really deployed 20 fabric WLCs to manage 20 fabric sites. What tasks/configuration during daily operation I would need to do on the WLCs one by one instead of using DNAC to accomplish centrally?