Sitting where they do, at the low-speed edge of the network, WAN access routers don’t attract nearly as much attention as their flashy high-density Gigabit brethren at the core. Yet the sheer volume of access routers deployed at most companies qualifies them for close scrutiny. Deploying the wrong access router to dozens or hundreds of locations can be a costly mistake.
While Cisco holds a leadership position in the network marketplace, it has a stranglehold on access routers. You can count more than a dozen firms that claim to compete with Cisco in the enterprise switching space, but you’ll find only a handful that are taking on Cisco in WAN access.
Of course, given the size of the market and the relative lack of competition, that is incentive enough for some companies to enter the market. In just the past 18 months or so for example, companies such as 3Com, Adtran, Enterasys Networks, Tasman and Vina (now Larscom) have marched through our labs to prove they’re bona fide competitors in this space.
In general, vendors vying for a piece of Cisco’s turf focus on several compelling points. Why Brand X? “Well, we are faster and less expensive than Cisco, we have greater functionality, and we’ll work with your existing Cisco network.”
While these statements can apply to both LAN and WAN infrastructure vendors, there are key differences that present unique challenges to WAN access gear vendors.
For starters, the traditional “faster” argument doesn’t apply the same way. Given that the T-1/E-1 WAN link (1.544M bit/sec or 2.048M bit/sec) is the gating factor, it is likely that every WAN router on earth is “fast enough” by this measure.
Thus, router vendors have taken to other methods to show the “power” of the platform. Much of our recent testing, for example, has focused on benchmarking the local Fast Ethernet routing performance of the access router.
This is not to imply that the average branch office needs wire-speed routing on its access router, but rather to illustrate the system has sufficient processing power and memory to provide a “future proof” platform – and thus becomes a good investment.
Cisco, for example, doesn’t even offer twin Fast Ethernet ports on its 1751. In 10M bit/sec tests, it delivers just under 30,000 packet/sec where some of the competition on Fast Ethernet tests reach close to 70,000 pps.
Interoperability, an issue on the LAN, is a bigger issue on the WAN. While on the LAN there is little direct interaction between Cisco and competing routers, serially connected WAN routers are tightly coupled in that each has to hold up its end of the chosen link protocol – PPP, High-Level Data Link Control or frame relay. Interest in this has prompted some vendors to have us conduct in-depth interoperability tests.
Competing vendors say Cisco offers an overpriced and underpowered platform that will require expensive (economically and logistically) upgrades – in effect forcing you to put a new router in place. The competition adds that their platform offers not only short-term but also long-term value by providing a future-proof platform.
I’d like to know what you think. Do you worry about having a future-proof WAN router? Do you have confidence that a non-Cisco router can be integrated into your network? What are your biggest access concerns?
Send me your thoughts on the subject. We’ll get them on Fusion, and I’ll try to distill them in a future column.
