[j-nsp] Miercom Competitive Performance Testing Results: Cisco ASR9000 vs Juniper MX960

[Kevin Oberman]

> Date: Fri, 25 Sep 2009 17:40:10 -0400 (EDT)
> From: Igor Gashinsky <igor at gashinsky.net>
> Sender: juniper-nsp-bounces at puck.nether.net
> 
> On Fri, 25 Sep 2009, Stefan Fouant wrote:
> 
> :: On Thu, Sep 24, 2009 at 9:30 AM, Richard A Steenbergen <ras at e-gerbil.net>wrote:
> :: 
> :: I thought this was the whole idea behind the introduction of the indirect
> :: next-hop.  Basically abstracting a level of recursion so that when
> :: underlying next-hop paths changed, all they needed to do was to do a KRT
> :: change for the indirect-next-hop pointer to the new path.  Any routes that
> :: referenced the indirect-next-hop then automatically utilized the new path
> :: via this mechanism, without having to do a KRT update for thousands of
> :: routes...
> :: 
> :: Am I missing something or did this behavior change at some point?
> 
> There are 2 levels where indirect next-hop can be used, one is on the RE 
> to speed up RIB convergence, and the resulting FIB computation, and the 
> other one on the PFE to speed up the actual FIB update itself. Juniper 
> implemented the one on the RE first, and that's been there forever, but 
> the PFE one is only available in the newer versions (and, can be enabled 
> via the "set routing-options forwarding-table indirect-next-hop" knob).
> 
> Configuring that knob will greatly speed up your actual convergence times. 

This looks like a big win, so I'm looking for a down-side. If there is
none, I would expect it to be the default.
-- 
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at es.net			Phone: +1 510 486-8634
Key fingerprint:059B 2DDF 031C 9BA3 14A4  EADA 927D EBB3 987B 3751