Proper ISP Backbone Design

[danarjamal]

Hello Dears,

I am searching for a way to disable spanning tree and use routing instead. and I was thinking about a combination of VXLAN, BGP and segment routing /or MPLS. 

I wanted to know which is the best practice for my scenario. let's say there are more than 300 VLANs in the network. does having 300 VLANs make the N93K switches overloaded? 

Kindly let me know your opinion and suggestion.

Thank you for your replies.

[balaji.bandi]

The context not clear, the subject says PROPER ISP design ( content related to nexus in DC environment ?)

I wanted to know which is the best practice for my scenario. let's say there are more than 300 VLANs in the network. does having 300 VLANs make the N93K switches overloaded? 

No it will not, again this is based on information.

do you have exiting network topology and also confirm is this DC environment or ISP design for customers?

MPLS/ VXLAN / MPBGP solves many issue, but depends on the design and where deploying.

[danarjamal]

Dear Balaji Bandi,

Sorry for not providing details, but I thought it was clear that this is not for DC design since I have mentioned ISP Backbone. This network is ISP Design For Customers.

Currently, our entire network is layer 2 between eNodeBs, customers until it reaches the ASR routers. and STP is blocking most of the links and it is not efficient, thus I want to disable STP and deploy segment routing with BGP, in that case, there will be L2VPN between eNodeBs and to the routers, and customers will be connected with L3VPN and L2VPN in some cases.

And since you have asked whether it is DC or ISP Design, I wonder if it is a good decision to use cisco N93K for my design or not? if not, what model do you prefer to support 200+ Gbps?

BR,

Danar