What would you do?

To say IT is a constant challenge is an understatement. Even the most scheduled of tasks can be a test of leadership. What sets good IT executives apart from great ones is how they identify and, ultimately, solve problems.

Three IT veterans - Rusty Bruns, CIO at Charleston Southern University in Charleston, S.C.; Jaime Man, IT director at Health Care Excel in Indianapolis; and Bill Randall, IT director at Red Robin Gourmet Burgers in Denver - tell us how they'd handle a few hypothetical, hairy situations. Weighing in on how their answers measure up is industry expert Johna Till Johnson, cofounder of Nemertes Research and Network World columnist.

CHALLENGE 1

You've been hired to overhaul an aging infrastructure: poorly performing mainframe and client/server applications, a routed network, a decade-old PBX, dial-up remote access and a frame-relay WAN with poor disaster recovery - generally 1980s technology everywhere you look. Over the years, business managers have built many workgroup LANs, so the company also owns thousands of servers on varying operating systems. What would you tackle first, and why?

Bruns: Overhauling an organization's technology infrastructure is a lot easier if you know who thinks they own what. Meetings or committees are a necessary evil. Users, managers, department heads must be pulled in to provide feedback on what they feel is the largest hurdle and what needs to be tackled first. Because this is a large organization, several committees will need to be formed. You don't have to agree with everyone, but to lead the group in the necessary direction. A sense of ownership by the organization in charge leads to a more productive transition, broad-based support and pushes disagreements more to the departmental level than against an IT leader and staff.

Once the political front has been addressed, it is then time to go to work. The mainframe is just a box that is running some form of ERP software. An ERP conversion is an 18-month to three-year project and while important, it is far more important to replace the mainframe with a server. Blade servers work very well, reduce space and power use, and allow for multiple servers to be used in one box. Because there are thousands of servers, it is a prudent purchase that provides for future expandability. Bringing the processing power/speed and reliability of a new server will have an immediate positive impact on productivity and boost morale as a major change has occurred. This process change can be operational within a few months. Emulation software can connect the new hardware technology to the old ERP software system.

Second, I would consolidate the servers and operating systems, again using blade technology. This allows for a large-scale, consolidated production server environment. The company would be able to use, depending on the age of the existing servers, approximately 80% to 90% fewer machines and make better use of underutilized processing power. With this in place, I would implement a storage-area network to take advantage of the benefits of centralizing storage. Where possible, I would standardize all desktops to one Windows-based operating system. This would ease the support burden for the help desk and support staff as well as the user. I would move the workgroup LANs into a centrally managed directory service-enabled network to reduce management costs and ease maintenance of user and computer accounts. It also would allow the standardization of user names and e-mail addresses.

Finally, I would upgrade the network. High-speed switches and routers on the WAN provide as much speed as possible to the infrastructure that was replaced or upgraded in steps one and two. One must ensure that the bandwidth is increased to take full advantage of the improvements. Without this final step, you're putting a garden hose on a fire hydrant.