www.opticallynetworked.com/features/article.php/3421291
By Kevin Reichard October 12, 2004 Normally, there's little in the router market that really shakes things up — a router is a router is a router, right? But this summer's release of the Carrier Routing System-1 router did introduce some intrigue and drama to what has been a flat and established marketplace. The Carrier Routing System-1, or CRS-1, is pitched by Cisco as heralding the next generation of core routers — high-end routers used by telcos and corporations to manage Internet data. And while much of what's contained in the CRS-1 is not anything new in the core-router world, it represents a significant change in how the networking world will approach core routers.
Now That's a Router While the growth of the Internet may have slowed a little over the last year, many expect it to explode again in the next few years as more users move to broadband and more stressful applications (VoIP, video on demand, messaging, videoconferencing) become more widely deployed. Because of this expected explosion in the IP world, telcos and corporations are expected to be aggressive about deploying new core routers. The needs of a telco and a corporations regarding core routers are not quite the same. Obviously, a Fortune 100 firm isn't going to spend a million dollars on a core router, and there aren't more than 10 telcos that would consider such an investment. And the aforementioned change in the core-router market currently applies to the high telco end of the market. But what happens at the high end of the market always trickles down to the rest of the market, and what is noteworthy about Cisco's new high-end CRS-1 router is technology that will also appeal to the rest of the core-router market: scalability. What's noteworthy about the CRS-1 is that it's designed to be expandable via a clustering structure. That is, new units can be added to an existing configurations without the need to install new routers. While expansion via clustering is not a new concept, it is a new feature for Cisco. It's also a desired feature for other core-router customers, who tire of replacing perfectly good core routers when expanding capacity. Replacing a router is a disruptive activity, and it's also a financially challenging one: Many times a core router needs to be replaced before it is financially depreciated. So the fact that Cisco is embracing a scalable core-router strategy is indeed big news.
It Moves Data ... Lots of Data (To further illustrate the difference between your average router and a core router: The CRS-1 occupies three refrigerator-sized units.) Cisco's endorsement of scalability legitimizes the practice for the rest of the core-router market, which has already embraced scalability. Newcomer Chiaro Networks garnered wide attention in the network world for its scalable approach to core routers. Juniper Networks, a strong player in the core-router market, has in theory supported clustering in recent years with its T-Series router, as has Avici, the number-three player in the market. The other feature associated with the CRS-1 that could change the core-router world is its reliability. Core routers, while expensive and geared toward the telco industry, have historically not been carrier grade, which carries its own special meaning in the telco world. (Basically, a component is considered carrier grade if it approaches a high level of reliability.) Because the routers were not carrier grade, telcos were forced into buying multiple core routers to assure redundancy. The CRS-1 is designed for carrier-grade usage, with several safeguards in place to prevent system crashes and failures. Still, the CRS-1 probably won't change the core-router world this year. The testing period for the unit is a little longer than usual, as Cisco is putting a new operating system through its paces. And when the CRS-1 first shipped, it did so without support for some pretty important features: IPv6 and MPLS (Multiprotocol Label Switching).
Pepping Up the Router World In fact, Cisco has seen its share of the core-router market decrease in the last two years, with rival Juniper moving from a 26 percent market share to a 32 percent share since 2002, according to Dell'Oro. Clearly, that loss of market share turned some heads at Cisco: The firm invested $500 million in development funding toward the project, which took four years to complete. The world of core routers isn't a glamorous one, despite the potential of having the contents of the Library of Congress sent to your desktop in under five seconds (of course, how to store it is your responsibility). However, with core-router technology set to change dramatically in the next few years, it's one that could quietly expand the capabilities of the Internet. |
|
|
|
|