Americas

  • United States
by Steve Taylor and Joanie Wexler

What about high availability in the branch office?

Opinion
Feb 20, 20032 mins
Networking

* Balancing budget with uptime

Last year, we covered the fact that the industry is moving toward more resilient IP networking. The IETF, for example, has been developing “nonstop routing” extensions to standard routing protocols such as Border Gateway Protocol.   For their part, equipment makers Alcatel, Avici, Cisco and Juniper, to name a few, have been introducing resilient router software features and hardware components in high-end products. These devices are targeted at carrier networks and the WAN edges of large enterprise sites.

Last year, we covered the fact that the industry is moving toward more resilient IP networking. The IETF, for example, has been developing “nonstop routing” extensions to standard routing protocols such as Border Gateway Protocol. 

For their part, equipment makers Alcatel, Avici, Cisco and Juniper, to name a few, have been introducing resilient router software features and hardware components in high-end products. These devices are targeted at carrier networks and the WAN edges of large enterprise sites.

Meanwhile, we’ve also reported increased competitive activity in the branch-office WAN access router space.  As organizations become increasingly distributed, they require more network connections in more places and that means additional capital and WAN service investments. These enterprises might not need certain expensive router features in all branch sites, however, so some suppliers now offer low-cost, basic-function IP WAN router alternatives.

Here lies the conundrum. To remain productive, distributed employees require reliable connectivity to corporate resources at headquarters. And as voice over IP joins the network, continuous uptime will be mission-critical. But it’s no secret that last-mile links and WAN access routers are the biggest contributors to remote-access downtime. And many organizations do not run redundant links and routers at remote sites for cost reasons.

You get the picture: How do you balance the requirements for highly available remote connectivity with budget limitations?

This is a question we will attempt to answer going forward with specifics from the various suppliers. The basic router software capabilities are there to switch a failed link or WAN router over to a backup. But the high-availability components of the big boxes – redundant route processors, graceful restart mechanisms, and sharing of state information between primary and backup routers – are just emerging for smaller sites.  We’ll look more closely at your options in upcoming newsletters.