Keep Your Network Up & Running
PROCESSOR | 02.01.08 | Easy Ways To Improve Uptime On The Network

Network uptime has been the Holy Grail of IT managers for years. It’s not hard to find horror stories about the thousands of dollars in lost sales for each minute a network goes down. It’s not hard to find tales of lost orders, appointment records, or some other transaction because a system went down, taking the data with it.

It should be clear that 100% uptime means no downtime. The terms don’t always make complete sense, however. If a network is up but one of its servers is overburdened and slows down network response times, is this downtime? Regardless of how you define uptime or downtime, there are many things that can be done to improve your company’s uptime.

 Verify, Test & Test Again

Many products can help to reduce threats to your system and maximize network uptime. Derek Wise, chairman and CTO of GNi (Global Netoptex Inc.; www.gni.com), a managed service provider, suggests that you take considerable care in planning any expansions and that you carefully audit your current structure. “In our opinion, the biggest thing where people fail is they implement new tools but fail to make sure they operate properly,” he says.

Wise believes in a “trust no one else” approach when it comes to choosing a service provider. “We never go into a data center before we test it out with live customers.” It’s sort of like the old political saying about the enemy—trust but verify. For the sake of your business, Wise suggests that you should know how things are working before actually betting your company on them.

“Trusting that a data center will give you an electrical diagram is not acceptable,” Wise says. “We pull cables; we break things; and we make sure that they’re not breakable before we turn them over to our clients. Once your systems pass your strict testing, you’ll be able to bring them live.”

 Control Access To Your Systems

Maintaining a maximal amount of uptime can be a serious challenge to system administrators. One of the reasons is that challenges to a system or network can come from many places. For example, Dan Clark, vice president of marketing at Lockdown Networks (www.lockdownnetworks.com), tells of a client who had “attacks on an ancient Telex machine that was sitting in a forgotten corner of their data center. They didn’t know it was there, and it wasn’t properly patched. Someone had gotten into their network through this Telex.”

Lockdown’s NAC products conducted a complete scan of the company’s network, detecting security vulnerabilities and reporting on the unknown Telex. “We deliver an appliance that assesses the devices on a network in two ways. First, there’s a network-based scan with tests going directly to the network devices, coming directly from our appliance. Second, you can install a software agent on a PC or a Mac. This agent then opens up a communication path to our product. It can do tests that we can’t do directly from our appliance,” Clark says. The Lockdown NAC is used “to assess all devices on a network—not just PCs—but anything on a network to assure that it is free from known vulnerabilities to attack,” he notes.

 Evaluate The Impact Of Change On Your Network

Kevin Kiely, president of GTC Networks (www.gtcnetworks.com), notes that you must carefully consider the impacts that changes will make before implementing them. “We offer a managed IP telephony environment, a VoIP infrastructure. Part of what we do is manage the whole process—we coordinate change management. One oversight that some people make is to make changes without any thought about the impact of the changes. The best thing (to maximize uptime) is planning and testing (before bringing anything live).”

“We heavily test. We make sure we have a great backup and recovery strategy in respect to our processes and systems. We have redundancy where we feel it’s necessary,” Kiely says.

Tom Coughlin, president of Coughlin Associates, a consulting firm and sponsor of storage-related conferences, also urges careful preparation to assure that any downtime is kept to a minimum. “One of the things you need is a good recovery plan,” he notes. “Plan for what could go wrong, create scenarios of situations that could happen, and let this guide you to put together a strategy for dealing with these problems.” Being prepared for disaster can help significantly reduce your downtime in the event of a system failure, a power failure, a successful attack on your network, or some other potential disaster. Coughlin says that an organization’s systems should be redundant and suggests that critical systems have failover capabilities. 

 

 

 

by Mark Brownstein