We have a customer who is keen to understand how Peplink can work in mixed deployments with legacy MPLS w/ CPE.
The have a core PE running multiple VRFs with overlapping subnets for their various customers. They wanted to know how they could run multi-tenancy with Peplink.
This is the customer’s diagram. I thought the best solution would be to run a Fusionhub instance in their core for each customer’s VRF, and use OSPF over separate VLANS (from the hypervisor).
I understand the best solution is to roll out Peplink at every customer site, fully mesh etc, but this is what the customer is asking. Is there a best practice or proven solution for this kind of scenario?
So either that or dedicated fusionhub per customer. Would need to know more about the core infrastructure / hypervisor options and levels of throughput / number of customers / remote sites to begin to advise which approach would be most ‘suitable’.
That’s great news. I assume FW 8 would support VRF on dedicated hardware too?, or is it specific to FusionHub?
I don’t have a lot of info on the existing core. Let’s assume it’s a Cisco PE with Cisco CPEs running either IPSec VPNs or leased line ethernet P2P. There is currently no provision of Hypervisor so this would be a ground up build specifically for running FusionHub (or Peplink hardware if that’s an option).
I’m based in UK and the customer is in Africa, throughput will be low (10mbit per site) and average number of sites per customer let’s say about 20. Total number of customers I will have to find out.
VRF has only been talked about for FusionHub so far.
I have done some testing with it and it seams to be great for eliminating the issue of overlapping subnets, especially when clients refuse to move from the default IP address!
No problem. Have done MPLS augmentation a bunch of times and the idea is all the same really - feel free to PM me with the diagrams if you have any and I’ll help pencil in where things would go.
