[c-nsp] Next step-up from 7206VXR

[Charles Sprickman]

On Feb 20, 2013, at 3:07 PM, Matthew Huff wrote:

> Once corporate networks decide they need IPv6 and they look at SHIM6, NPTv6, and multihoming and see how many apps break out of the box, they are going to request /48 Provider Independent (PI) spaces  and advertise them via BGP. Many if not most have been burned by PHB signing telco deals without any buy-in, so they aren't going to want to have to renumber every time some contract is signed/broken. Easy IPv6 corporate prefix renumbering is a myth. Too many one-off embedded devices that barely speak IPv6 and monitoring systems that aren't dynamic and other issues (ACL rewrites, etc...)
> 
> The explosion in ipv6 prefixes is coming....
> 
> Unless you need the hardware switching capabilities of the 7600/6500 (microbursts, etc), why not use the software routers like the ASR 1000/9000 that have no fixed TCAM limits?

Going by the parallel thread running here on the ASR1000 series, you actually do have a fixed limit on the 1001 and 1002-X series (and a limit based on which ESP you put in the more modular chassis).

https://puck.nether.net/pipermail/cisco-nsp/2013-February/089548.html

While Mr. McBride does not call out actual numbers there, so far that's the most information I've seen about the 1002-X.  On paper at least, the fixed config ASR1Ks have the same 1M route limit in the FIB (or whatever it may be called on that platform) as the various sups being discussed in this thread.  It looks like you have to move to the 1004 and at a minimum drop in the ESP-20 to get 4M routes in the FIB (and looking at some quotes, putting an RP2 and an ESP-20 in a 1004 chassis is about double the street price of the 1002-X).

I can see why Cisco is gimping the fixed-config models, but it's quite frustrating - they must know the stuff is going to obsolete itself if used in an SP environment.

You can peruse the whole mind-bending thread:

https://puck.nether.net/pipermail/cisco-nsp/2013-February/089302.html

Charles

> 
> 
> 
> 
> -----Original Message-----
> From: cisco-nsp-bounces at puck.nether.net [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Seth Mattinen
> Sent: Wednesday, February 20, 2013 11:20 AM
> To: cisco-nsp at puck.nether.net
> Subject: Re: [c-nsp] Next step-up from 7206VXR
> 
> On 2/20/13 6:13 AM, Mikael Abrahamsson wrote:
>> On Wed, 20 Feb 2013, Mack McBride wrote:
>> 
>>> At 768k you are effectively limiting your IPv6 table to 128k (you
>>> can't really go more than that if you expect to use IPv6). I recommend
>>> a 640k/192k split.
>> 
>> Well, I believe IPv4 will hit 640k before IPv6 will hit 128k, so I'd
>> recommend 768k/128k instead.
>> 
> 
> That's what I think too, unless there are too many unchecked idiots that
> decide to deaggregate their /30-32 into /48s.
> 
> ~Seth
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
> 
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/