[c-nsp] 12410 or 7690-S for BGP transits

[Deepak Jain]

YMMV, but I find that equipment that is heavily redundant (internally) 
is sometimes a waste compared to the additional cost of just deploying 
an additional chassis. If you are buying dual supervisors, you might be 
saving the second licensing costs on the software, but other than that a 
chassis and power supply are pretty cheap. Another way to look at is... 
you are spending 160% to be *mostly* redundant where for 195% (assuming 
a discount) of the cost you can be absolutely, completely, physically 
redundant. (Unless space is a concern).

Whether its for IXP use, or whatever... a 6500 or 7600 do really well 
with lots of and lots, and lots, and lots of low latency ethernet 
devices connected to them and 10Gb/s of thruput is nothing. If you plan 
to run your links at a high usage level or have long distances between 
end points, consider some more expensive (SIP600 or whatever GSR 
comparable) as needed.

For your customers who need "high availability" dual home them to two 
identical pieces of equipment... you can service as needed and not 
pray/hope that SSO works as advertised.

Deepak Jain
AiNET

William Jackson wrote:
> Its more a case of IOS features and platform capabilities than port
> densities and costs, as both platforms can have high performance and
> port densities.
> 
> Like I mentioned before, I would like the uptime to be as good as
> possible, with regards to software upgrades and self healing in event of
> bugs etc.
> 
> We are using BGP and OSPF on our current boxes so would expect to keep
> the same protocols.
> 
> thanks
> 
> 
> -----Original Message-----
> From: Phil Bedard [mailto:philxor at gmail.com] 
> Sent: 31 January 2008 15:55
> To: William Jackson
> Cc: cisco-nsp at puck.nether.net
> Subject: Re: [c-nsp] 12410 or 7690-S for BGP transits
> 
> Generally for BGP peering you don't need all the fancy features that  
> the SIP-600 gives you, it's complete overkill.  You are better off  
> just getting a 7600 with normal LAN linecards, otherwise your per-port  
> cost is going to be enormous.
> 
> Phil
> 
> On Jan 31, 2008, at 4:00 AM, William Jackson wrote:
> 
>> Hi all
>>
>>
>>
>> I am looking for a platform for the BGP peering edge of our network,
>> this will have multiple full BGP tables and carry all our transit
>> traffic.
>>
>>
>>
>> We expect to have between 5-10 gbps of initial traffic capacity and
>> expect to grow over the net couple of years.
>>
>> We are all Ethernet based shop.
>>
>>
>>
>> I am considering two options:
>>
>>
>>
>> Option 1:
>>
>> 12410 ( with 2 * PRP2 and SIP601 line cards ) running IOS XR
>>
>>
>>
>> 7609-S ( with 2 * RSP720-3CXL and Sip-600 line cards )
>>
>>
>>
>> The requirements are for:
>>
>>
>>
>> Carrier grade equipment
>>
>> High availability
>>
>> ISSU, NSF and SSO
>>
>>
>>
>> We require these devices to have as many high availability features as
>> possible to provide as much of a "continuous operation" scenario as
>> possible.
>>
>>
>>
>> I heard that the new 720-3CXL has a very fast 100msec switchover  
>> whilst
>> the 12XXX was a lot slower to converge in a failure scenario??
>>
>>
>>
>> Any recommendations and pointers of caveats etc to consider?
>>
>>
>>
>> Many thanks
>>
>>
>>
>> _______________________________________________
>> 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/
> 
>