[j-nsp] Advice on a 100Gbps+ environment

[Mikael Abrahamsson]

On Tue, 2 Jul 2013, Morgan McLean wrote:

> of row switches. This leaves me with hoping that OSPF ECMP works well
> enough to push these kinds of traffic levels over a bunch of 10GE links,

OSPF is control plane, it'll set up the ECMP and tell the forwarding plane 
about it. On some platforms, you'll get the same load sharing behaviour 
regardless if it's L2 (LAG setup with perhaps LACP) or L3 (ECMP setup by 
OSPF).

You never said what kind of traffic this is supposed to handle. If it's a 
lot of parallell TCP or UDP sessions, then you'll most likely be fine with 
LAG/ECMP.

> Also...any tips on handling all of the provider connections? I don't 
> know if they will be giving us lags, which means I'll be potentially 
> running 10-15 BGP sessions per MX, which is a lot of routes. They're 
> running the RE-S-1800 quad core 16GB, but should I opt for partial or 
> just default at that point? Its possible all of the links will be from 
> the same provider.

I'd suggest either LAG or ECMP (in case of ECMP, you need to do a static 
host route to other end on all components, and then do the bgp session to 
this /32 or /128 with ebgp multihop enabled). Then you only need single 
session per provider.

-- 
Mikael Abrahamsson    email: swmike at swm.pp.se