Alcatel Unveils New Routing Technology
Alcatel says it's found a way to ensure that core routers stay up and running properly all the time, without the delays associated with router reconfiguration. Alcatel also says its new technique fosters scaleability, another focal point for carriers.
Alcatel's solution, dubbed ACEIS (for Alcatel Carrier Environment Internet System), improves performance by eliminating delays that happen when routers try to communicate with one another in order to reconnect around a broken link, the vendor claims. To eliminate the floods of messages that result, ACEIS streamlines the performance of protocols such as BGP (Border Gateway Protocol), OSPF (Open Shortest Path First), and IS-IS (Intermediate System to Intermediate System) during router recovery, Alcatel says (see Alcatel Debuts Non-Stop Routing).
Alcatel has applied for patents on ACEIS, which it says will be added to its product line sometime this year, starting with its 7770 Routing Core Platform and its 7670 Routing Switch Platform. The vendor plans to demonstrate ACEIS at the Supercomm 2002 tradeshow in Atlanta.
Alcatel also plans to subject the technique to testing with BTexact Technologies, the research and technology business of British Telecom (BT) (NYSE: BTY), at an unspecified date this year (see BTexact Preps Router Test). "Instead of saying we have betas with customers we can't reveal, we'll turn to a third-party lab for validation," says Ben Crosby, senior technologist for IP strategy and solutions in Alcatel's broadband networking division.
Alcatel's ACEIS is just the latest in a slew of efforts to assuage carrier fears about IP that appear to be surfacing as carriers contemplate more dependence on the use of the protocol in their networks.
"There is still a lingering perception in the carrier environment, rightly or wrongly, that IP routers are not as robust as voice switching gear," says David Newman, president of Network Test Inc.
Alcatel uses this perception as a starting point. "Our information shows that router-related problems cause over 50 percent of downtime in wide-area networks," says Vinay Rathore, director of strategic marketing for Alcatel's Broadband Networks division.
ATM switch vendors that compete directly with Alcatel have their own claims to be "carrier class." These vendors include Lucent Technologies Inc. (NYSE: LU) and Nortel Networks Corp. (NYSE/Toronto: NT), which, like Alcatel, are intent on adding routing capabilities to multiservice switches based on an ATM platform (see The Great ATM Switch Blitz). It's clearly in these vendors' interest to claim solutions that are more reliable than those of router vendors such as Cisco Systems Inc. (Nasdaq: CSCO).
For its part, Cisco claims to have carrier-grade reliability in its products now and is working on more. It says the ability to forward packets nonstop in the face of failed segments is a key differentiator for the Cisco 12000 Series Internet Router.
Cisco rival Juniper Networks Inc. (Nasdaq: JNPR) also is expected to release some reliability augmentations with its much anticipated next-generation core router (see Juniper Mum on Core Router... and Report: Core Router Market Falls 22%).
Other vendors, such as protocol software supplier NetPlane Systems Inc., have products they say address the issues involved. NetPlane recently announced an OSPF stack it says is fully redundant, cutting down on the reconfiguration messages that occur in networking gear in the face of a broken segment (see NetPlane Ships OSPF Solution).
Besides these proprietary efforts, several projects have been undertaken within the Internet Engineering Task Force (IETF) to make IP routing protocols more efficient in general. These include the "Hitless OSPF Restart" working group, also known as the Non Stop Forwarding (NSF) group, which has significant participation by Cisco and Sycamore Networks Inc. (Nasdaq: SCMR).
For its part, Alcatel claims that most of these efforts, particularly Cisco's and those in the IETF, merely slap a bandage on the problem. Even if they result in nonstop forwarding, performance can still suffer when segments break. What's needed, Alcatel says, is technology (like ACEIS) that changes the way IP works in the hardware itself. "If you deal with the issues that cause forwarding delays, you won't have to extend the protocols in order to minimize outages," says Ben Crosby.
Alcatel is being cagey about whether the ACEIS design involves the use of ASICs (application specific integrated circuits). "I'd like to tell you about the design, but that's information we haven't disclosed," says Crosby. "I can say that the design is not dependent on the use of ASICs."
Analysts say there's merit in both the proprietary and would-be standards for improving IP routing reliability, and that the market will determine the best solution. "NSF may be the only way to implement this type of capability in Cisco's installed base," says Mark Seery, program director at RHK Inc.
"If NSF proves to be not as good as Alcatel's approach, well, at least Cisco customers are better off than what they are today, especially with respect to software upgrades. If NSF proves to be as good as Alcatel's approach, then that of course diminishes the value of what Alcatel is doing." Seery says he "looks forward to the marketplace having access to both approaches and kicking the tires on each." And he's confident the upshot will be greater reliability in carrier facilities.
Some question whether ACEIS and other efforts are not serving more of a marketing than a technological need. David Newman of Network Test says IP routing gear "could be more reliable," but he points to the extensive redundancy that vendors of switches and routers already have in their gear.
"They do NEBS [Network Equipment Building Systems] certification... the same certification that supposedly more reliable switches undergo," he says. "It's the same test. I don't think it's useful to distinguish between the two classes of devices."
— Mary Jander, Senior Editor, Light Reading