Are IP networks getting more resilient?

Packet data networks must meet or surpass the so-called " five nines " availability levels of the public telco network before enterprises will fully embrace converged IP networks for all their mission-critical applications. Customers today are accustomed to having no more than 15 seconds of downtime per year on their circuit-switched voice networks. So they are unlikely to tolerate iffy voice quality on IP networks in which router reconvergence after a failure takes many minutes.

Strides are indeed being made in routers and switches on the high-availability front. Most vendors have begun by building redundant hardware components - CPU, line cards, power supplies, and fans - into their systems. Meanwhile, on the software side, the ability for routers to maintain state information in the event that one component does have to fail over to a backup (or one router has to fail over to another) will hasten router reconvergence and, thus, network recovery times. The faster the recovery time, the less impact there is on application performance.

Cisco, for example, recently announced enhancements to its IOS router software, collectively called " Globally Resilient IP, " that will ship throughout the year beginning in June. The " crown jewel " enhancement - a feature called Nonstop Forwarding with Stateful Switchover - protects against route processor failures, says Charles Goldberg, a Cisco product line manager. It does this by enabling a standby route processor to maintain state information of the primary route processor (such as all the Layer 2 connectivity protocol information) so that during a failover, the backup can pick up right where the primary left off.

This feature contrasts with Cisco Route Processor Redundancy Plus, used heretofore in Cisco gear, which enables a standby route processor to take over, but does not continually update state information between the primary and back-up processors. Rebuilding the state information in the backup, then, consumes time and degrades session quality. According to lab tests conducted by Miercom, Princeton Junction, N.J., in April, for example, the new capability reduced the mean time to recovery of the Cisco 12000 Series Internet Router (the company's high-end, 10-gigabit-capable router) from 37 to zero seconds with zero packet loss under high-stress situations.

What does all this mean to user experiences? We'll take a look at that next time.

Cisco Globally Resilient IP
 

Steven Taylor, consultant and broadband packet evangelist, and Joanie Wexler, an independent networking technology editor and writer, team up to bring you this analysis and commentary. Taylor specializes in education and market analysis, and Wexler adds incisive reporting and research. For more detailed information on most of the topics discussed in this newsletter, connect to www.webtorials.com, the first Web site dedicated exclusively to market studies and technology tutorials in the Broadband Packet areas of Frame Relay, ATM, and IP.

Feedback and additional topic ideas are welcome. Please contact taylor@webtorials.com or joanie@jwexler.com.

Frame Relay archive
Past newsletters.