Controllerless Networks

I’m going to design an Aruba Wi-Fi 6 solution on a new customer and I have a doubt about the pros and cons of Instant solution compared to Mobility controller solution especially related to scalability.

My customer has only one site now (site A) but in the near future, probably in 1 year, will build a new much bigger site (site B) and he wants to cover the entire area with Wi-Fi as well. Thinking ahead I have two solutions (let’s say in the new site 20 AP will be required):

1. ALL-INSTANT SOLUTION à In the site A I will install 10 500 series APs in Instant mode with cloud license, then after the site B will be ready I create a new Instant cluster with 20 500 series APs and cloud license as well.
2. MOBILITY CONTROLLER SOLUTION à In the site A I will install 10 500 series APs with a mobility controller virtual appliance, then after the site B will be ready I will add the 20 500 series AP to the mobility controller in the site A.

In the 2nd solution can I manage the APs in the site B from the controller in site A if two sites are connected by high speed fiber? Could it be the 2nd solution an accettable architecture technically speaking?

Thanks for the answers

For the number of APs you are designing for, i would go with Aruba Instant.

it has almost all the features of Controller based solution with the simplicity of deployment/mgmt. It can scale up to 128x APs with multiple layers of redundancy and Commercially you don't need to get 2x controllers (for redundancy ) and its associated licenses.

The most important part of the Instant Architecture is that it has Distributed forwarding and Crypto plane which means all the traffic is bridged out of the local Ethernet port of the Instant AP. rather than being GRE tunneled to the controller.

Lastly make sure you also have Aruba Central to manage the two sites central from one portal.

Hi ariyap and thanks for you answer.

when you say "all the traffic is bridged out of the local Ethernet port of the Instant AP" you mean that all the traffic doesn't flow through the VC (as in the controlled mode does) but it flow directly from the APs to the local switched network (by using the local eth port of VC) so the VC doesn't represent a single point of failure?

that is correct.