[f-nsp] FES-XE Performance Q
Jason J. W. Williams
jasonjwwilliams at gmail.com
Tue Dec 9 18:40:38 EST 2008
Hi Niels,
Thank you for the detailed response. The FES-XE can hold 1M entries in
its RIB and 500K in its FIB, so I think from that point of view it
will barely let us cram in full routes from three peers. Otherwise, I
think we have to consider an MLX-4 or a Juniper M7i.
-J
>
> Message: 5
> Date: Tue, 9 Dec 2008 16:34:39 +0100
> From: Niels Bakker <niels=foundry-nsp at bakker.net>
> Subject: Re: [f-nsp] FES-XE Performance Q
> To: foundry-nsp at puck.nether.net
> Message-ID: <20081209153439.GT78345 at burnout.tpb.net>
> Content-Type: text/plain; charset=us-ascii; format=flowed
>
> * jasonjwwilliams at gmail.com (Jason J. W. Williams) [Tue 09 Dec 2008, 08:17 CET]:
>> I have somewhat of a general performance question regarding the L3
>> performance of the FES-XE line. Foundry is claiming 80 MPPS, but I
>> assumed that was a marketing number that would degrade with smaller
>> packet sizes. My rep is telling me that that is the performance even
>> assuming 64-byte packets. Is that everyone's experience with Foundry
>> gear? Just having trouble believing it because at 64-byte packets that
>> would be CRS-1 performance.
>
> Pretty much, yes. Of course with CRS-1 and IOS-XR you get an actual
> chassis-based router with router features such as fully-featured MPLS
> and enough memory to carry multiple full routing tables with room for
> growth and 'continue' in route-maps* and linecards and upgradability and
> a forklift to install it.
>
> I'm glad that Foundry tends to publish Mpps numbers for 64 byte frames
> and not do the "double all numbers and call it full-duplex" marketing
> stunt that Cisco is so often guilty of.
>
>
> -- Niels.
>
> *) Actually currently broken in IOS-XR. :)
>
> --
>
>
> ------------------------------
>
> _______________________________________________
> foundry-nsp mailing list
> foundry-nsp at puck.nether.net
> http://puck.nether.net/mailman/listinfo/foundry-nsp
>
> End of foundry-nsp Digest, Vol 71, Issue 6
> ******************************************
>
More information about the foundry-nsp
mailing list