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

[Eric Van Tol]

> -----Original Message-----
> From: juniper-nsp-bounces at puck.nether.net [mailto:juniper-nsp-
> bounces at puck.nether.net] On Behalf Of Richard A Steenbergen
> Sent: Thursday, September 24, 2009 4:12 AM
> To: Jeff Tantsura
> Cc: juniper-nsp at puck.nether.net
> Subject: Re: [j-nsp] Miercom Competitive Performance Testing Results:
> Cisco ASR9000 vs Juniper MX960
> 
> On Wed, Sep 23, 2009 at 10:46:01PM -0700, Jeff Tantsura wrote:
> > Hi,
> >
> > Would be interesting to see Juniper's reaction on following report:
> >
> > http://www.miercom.com/dl.html?fid=20090827&type=report
> 
> These reports are total frauds, Cisco pays Miercom to produce them and
> to make sure the results always come out in their favor. They typically
> find edge cases to focus on, misconfigure the competition boxes, provide
> no way to verify their claims, and make sure to never feature any test
> in which the Cisco box fails. The Miercom 7600 vs M7i report for example
> was a work of pure comedy, if not for the thought that some poor sap
> might have actually believed a word of it. The FUD machine has been
> working overtime against the MX lately too, because it is an first rate
> box and the competition has absolutely nothing to compete against it
> technically.
> 
> --
> Richard A Steenbergen <ras at e-gerbil.net>       http://www.e-gerbil.net/ras
> GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)

If I remember correctly, they configured things that caused some sort of memory search cycle to be exhausted, which is something that is not possible in a TCAM-based solution.  In addition, they used "cheap" LAN modules to make it look as though the 7600 was "low cost", even though they fail to mention that those LAN modules don't support the same features that the M10i GE ports did.  You would have to purchase OSM/SIP modules to support those same features, which raises the cost considerably.  There was also something having to do with the specific number of ACL entries, which makes me question why in this new report, do they choose specifically 2010 and 4020 VPNs.  Not sure if it has any significance, but it sounds hinky to me.

-evt