Tired of band-aiding recurring information technology problems at the failure point, Hannaford Bros. is turning to a network management software to spot IT problems before they result in down time.

The Scarborough, Maine-based supermarket chain, which operates 154 stores in New England and the Carolinas, was prompted to look into the management system when it noticed that its network was not polling systems fast enough to detect problems. "We saw many systems would be completely down by the time our support desk received the call from the user," says Rick Lemelin, senior systems management analyst, Hannaford. "The new solution lets us see and fix a problem before there is down time, so no one is interrupted."

Partnering with Tivoli Systems, an IBM-owned provider of IT management systems, Hannaford is implementing the vendor's Enterprise suite, which proactively monitors its best-of-breed IT environment from a central point.

"For each day a system is down, a fraction of customers will drop orders down and walk out of the store," says Shlomi Harif, Tivoli's director, pervasive business technologies. "That could mean a lot of lost sales."

Adopting ATM: Hannaford began the upgrade of its monitoring system by switching to an asynchronous transfer mode (ATM) network from a satellite communications system. ATM affords much faster reporting of outages.

The chain then got started in network management with the installation of Tivoli's NetView, an application that monitors the hardware that makes up its network, including all hubs, routers and servers.

Hannaford then installed Tivoli's Management Agent, a two-way communications application, on each desktop, laptop and Unix server. "Instead of polling and getting reports on 2,200 devices, we install the agents on 160 interfaces and can poll the entire network in 15 seconds," Lemelin explains.

Hannaford just added another 160 agents to support its open NT platform. "Both [the NT and Unix agents] will run simultaneously, so we do not have any silos for any systems that may run on different platforms," Lemelin explains. "Whether the failure is on NT or Unix, our help desk will see it."

To better track performance and availability of servers throughout the chain, Hannaford is also using a Distributed Monitoring module that enables the support staff to set and change parameters to monitor multiple remote systems.

Taking action: When systems do not meet parameters and begin malfunctioning, Tivoli's Manager for Network Connectivity pinpoints the cause of the problem, then records the "symptoms" in a centralized database. "Alarms" are sent to Tivoli's Enterprise Console (TEC), or centralized mission control, which uses artificial intelligence to formulate a response.

In some cases, automated responses, such as restarting a system, can be programmed into TEC. Hannaford also plans to add a function that would allow the system to automatically page a systems administrator to respond to outages. "This is our gateway to problem management," Lemelin says. The complete system went live last October. Thomas Witwicki, director, technology services for Hannaford, declined to comment on results, saying it is still designing metrics to evaluate the system, though he noted that the company has forecast has a five-year return on investment for the system.

Hannaford also plans to implement Tivoli's Software Management module by the second quarter, according to Witwicki. The application will automate distribution of new software and version updates from a central location to more than 2,200 desktops residing at headquarters and the store locations.—Deena M. Amato-McCoy