How to read "Routing Performce" data and apply to router selections

[toshi-esumi]

We often refer to the white paper from Cisco below to decide which type of router to deploy for different situations.

http://www.cisco.com/web/partners/downloads/765/tools/quickreference/routerperformance.pdf

However, at last for me, it's not clear that the numbers on the paper is based on one-way data transmission between port-A and port-B, or bi-directional simultaneous transmit&receive between two full-duplex ports. If the former, should we divide the numbers by 2 when we consider full-duplex performance and choose a router accordingly?

And/or if these kinds of performance tests have some standards, please tell me where to look for them.

Thanks,

Toshi

[pascalfr0]

Numbers are given for one way throughput.

It's just pps * 64 * 8.

64 bytes packet is worst case scenario, not realistic (in real life, medium packet size should be around 350 bytes on most networks with mix of traffic close to IMIX). In the other hand, every feature that will be activated on the router will get its pps performance down (QS, ACL, NAT, tunneling).

You should talk to Cisco staff about what's your traffic profile and what feature you want to use. They have scaling tables for the most common features based on internal testing.

[toshi-esumi]

Thanks for the additional info. It's often unfortunate that we couldn't spec out in detail what it would have or don't have enough time to do when we sell, other than knowing just WAN circuit bandwidth. So we need to have some guide line ourselves including some margin of errors.

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

When using that document, the performance would be the aggregate for all traffic being forwarded.  For symmetrical duplex, dividing by 2 would be correct, but allowing for all the other things a router needs to do, and often does, and leaving some excess capacity, dividing by 4 puts you in the ballpark.

I've attached a Cisco document that better explains real-world performance of many of their ISRs.

[toshi-esumi]

Thanks a lot for sharing this well-written document. I searched through but couldn't find like this. I'll read through to set our guide line.