nanog mailing list archives

Re: Is Cisco equpiment de facto for you?


From: "Justin M. Streiner" <streiner () cluebyfour org>
Date: Mon, 10 Jan 2011 07:02:17 -0500 (EST)

On Mon, 10 Jan 2011, Brandon Kim wrote:

Would you say that it's fair to say that if you are serious at all about being a service provider that your core equipment is Cisco based?

I would not necessarily say that. Granted, most of the places I've worked are Cisco shops to a large extent, that does not mean it is the only solution to many problems. There are product spaces where Cisco has a very well established presence (routers, switches, remote access, wireless, DWDM, IP telephony, etc), but there are other players in those spaces, in additional to spaces where Cisco does not have as large of a presence.

There are many people who will give 'buy Cisco' as a default answer to many networking needs, much the same as there are meny people who will give 'buy Microsoft' or 'buy Oracle' for software/database needs. There are other solutions to those needs. If you see a piece of gear from a new vendor, don't be afraid to contact them to see if get their sales team in to give you a presentation or take a box for a test drive. Ideally, you also have acess to some type of lab or non-production environment where you can try out equipment without putting 'live' data at risk.

In most shops I've worked in, the final decision comes down to:
1. cost
2. performance/reliability
3. support
4. scalability (read: investment protection, speaking back to point 1)
5. interoperability
6. security
7. environmental factors (rack space, power, cooling, etc)

Pretty much all of the subsequent points ties back to point 1 in some way.

Network devices are tools designed to do one or more jobs. The job you're trying to do determines the tools you use, and how you use them.

jms

PS: I take test results from Tolly/Gartner/Burton/etc with a grain of salt. When a vendor performs well in a bake-off, they will proudly trumpet that fact. When they don't they will usually claim that either the box they tested was broken, or the testing methodology was flawed in some way :)


Current thread: