The recent Cisco negative marketing campaign against Juniper surprised and disappointed me, see the following URL for info http://www.overpromisesunderdelivers.net/pdfs/Why_Cisco_Not_Juniper.pdf
I remember being amazed by the capability of Cisco equipment in 1997, when I started building ISP infrastructure. You could create blisteringly fast Internet backbone networks using 7200 and 7500 routers. The marketing was slick and there wasn’t really anything else on the market that you would want to build your network with.
A few years later the growth of the Internet accelerated to the point that we started to work around the limitations of core routers and Internet scalability using techniques like route flap damping. It’s been a while since I built Internet backbone networks, I understand that route flap damping isn’t used any more as the solution became more cumbersome than the problem. Now Internet backbone routers are fairly chunky beasts with control and forwarding plane separation making RFD redundant.
The lack of control and forwarding plane separation seemed to be the start of the demise of the dominance that Cisco enjoyed in the backbone router market. The emergence of the Juniper M40 router circa 2000 marked the start of the erosion. Large service providers were struggling to cope with the demands of Internet growth.
Traffic was growing rapidly, an Internet backbone year was equated to being 90 days. Core networks were having to plan network capacity changes that would normally last for a years worth of growth, Internet change was so fast that 90 days later you were planning the next upgrade. Exciting and fun days, Juniper came to the rescue of the largest providers, offering great, yet flawed routers (the M40 had active backplane components), that enabled service providers to cope with the growth of the Internet without sacrificing performance.
Juniper engaged with potential customers, educating them about the benefits of their router architecture without drawing heavily on comparison with other vendor equipment. This marked the start of my journey in understanding how hardware bolted together and the benefits that can be gained by hardware design philosophy. Juniper’s approach was engineering led with very little marketing hype.
After Juniper gained some market share, Cisco produced some great hardware (perhaps a little late to market) but they didn’t resort to any cheap shot marketing stunts. The recent campaign is very disappointing, I would like to see educate customers on why their hardware architecture is great compared to the competition.
Come on Cisco, supply us with information that gives us confidence in the products you produce and give us a an alternative to Juniper for our core routing needs. Until you start doing this, engineering led service providers will continue to use Juniper in the core.