New Tools Ease the Burdens in Network Management

By Heather HayesAs networks grow and become increasingly complex and critical in this age of electronic business, both government and business organizations face a widening gap between their technology needs and the availability of qualified personnel. Nowhere is this disparity more apparent than in the area of network management."Traditional data net monitoring tools have forced network operating center staff not only to react to problems rather than prevent problems, but also to waste countless hours simply tracking down the cause," said Susan Almeida, a leading consultant and co-founder of Network Strategy Partners LLC, a Boston management consultant firm for the networking industry.To deal with the problem, more companies specializing in network management have begun adding intelligence and usability features to their existing management platforms to help ease the work load. Meanwhile, several small start-up firms have developed Web-based middleware products to integrate, correlate and aggregate the performance and event information generated by those management applications.The combination of the two product categories will go a long way toward making it easier not just for in-house staff, but also for systems integrators looking to support and manage networks for federal agencies. "All of these tools are welcome additions for us," said Denis Brown, vice president and general manager for Air Force Infrastructure Systems for Litton-PRC Inc., a systems integrator in McLean, Va. "We're all just looking for models where we can be more efficient in our utilization of people and use technology more effectively, so that our most qualified people can concentrate on other more critical tasks."The two categories of products are being driven by many market factors: the growing importance of the network as a critical piece of the infrastructure for the successful delivery of e-services; the scarcity of qualified IT workers; and, thanks to the increasing reliance on service-level agreements, the need for quick, efficient and effective reporting methods.For both in-house organizations and service providers, the new tools will provide immediate and visible benefits, especially when used in tandem, according to Rich Ptak, vice president of systems and applications management for the Hurwitz Group, an IT industry analytical firm in Boston."Organizations will be able to do more of the management tasks and take care of larger networks, more nodes and more devices using less sophisticated workers and without having to add new ones," he said. "With these tools, when you have a router go down or some other issue with the network, you don't have to call in your most senior staff person." A less experienced person can take the corrective action, Ptak said, adding: "You'll still want to have an experienced network management individual there, but at least you can get some of the problems and troubleshooting started."XXXSPLITXXX-The trend toward tools that are easier to use begins with the platforms. Most major vendors, including Tivoli Systems Inc., Computer Associates International Inc. and Hewlett-Packard Co., are building intelligence into their products so that information gets processed more effectively, and the user performs less interpretation. "When an event occurs, the platform would give you the information that managers would immediately need, but more importantly it would provide context information," said Ptak.Take, for example, NetView 6, the latest version of the network management system put out by Austin, Texas-based Tivoli. It features these enhancements:• A router fallback relations function. "Previously, network management tools would notice that a large part of the network had disappeared, and immediately conclude that everything, everywhere is failing," said Jim Carey, director of NetView Applications Development for Tivoli. "Instead, this function zeroes in on the specific device that's failing and indicates which systems beyond that device are affected by the failure, so the network administrator can immediately take action against the failure point and then gain an understanding of what the impact is on the overall network."•SmartSets. This component allows organizations to dynamically group systems, network devices and business systems by their like attributes, their similarity within the network (as in all routers) or their business purpose, such as anything related to an external Web site. As a result, according to Carey, organizations can handle data collection, polling and other down-and-dirty tasks of a network management tool in one fell swoop. "Do it once, and forget about it," he said.• A location-sensitive topology function. This allows administrators to lay out the network map according to the way they individually view the network. Although most network maps are designed according to the TCP/IP information used to route traffic, most administrators take their cues as well from where the system is located and what it does. The new function lets them alter the look of the map, so it is more in line with their own thinking and the organization's specific needs.The overall result, Carey said, is that "it will be much easier to recognize the source of the problem, so that an operator's attention will be drawn immediately to the source of the problem in your network. For the organization, that means shorter time to understand what problems are occurring, and shorter time to resolve those problems."The market for easy-to-use features is so strong, in fact, that even relatively young companies are taking a crack at the game. For example, Entuity, a 3-year-old firm based in New York, recently released a product called Eye of the Storm. It analyzes mission-critical networks with drill-down and issue-prioritization capabilities, giving IT departments the ability to anticipate and address performance degradation before the business is affected."It interrogates all the different components on the network and maintains a real-time monitor of its current performance, as well as a history of its performance," said Barbara Harrison, vice president of marketing for Entuity. "All that information is put into a business context, so it tells you not only the very specific network address where there's a problem occurring, but what the actual device is that's connected there and what the overall impact will be. All of that helps the operations team more easily prioritize what they work on."XXXSPLITXXX-Unfortunately, most organizations are outfitted with not just one, but numerous independent management tools, a fact that leaves staff faced with several screens, each one displaying different information from a different tool. "They have to manually aggregate and correlate all that information and try to interpret what's happening, or launch another tool to do that," Ptak said. "Sometimes they don't even have enough screens, so they have to unplug from one processor and plug into another. Meanwhile, they're copying down data by hand onto yellow sticky pads and going from terminal to terminal. It can be really chaotic and is extremely inefficient."Several vendors, however, recently have come to the rescue with new middleware products capable of translating the different tools, correlating and aggregating the information and displaying it all on a single management console.These vendors include Edge Technologies of Fairfax, Va., which just released its Web-based middleware product, En Portal, in May; and McLean-based Managed Objects, which is just beginning to market its Web-based integration platform, Formula, to the government. Both products allow network administrators to perform easily predictive analysis, trend analysis and other preventative measures, as well as quickly and effectively determine where problems lie on the network.While the desire for speed, efficiency and ease of use in network operations centers is driving the push for middleware, other more market-specific factors also come into play. Larry Chang, president and chief executive officer of Edge Technologies, said a major factor is the need for new user constituent groups to view networking data."A vice president of sales might want to see what kind of network latency his salespeople are experiencing," he said, noting that as a role-centric product, En Portal can be programmed to include only the information relevant to each user. "Rather than create one standard, normalized view, we let them aggregate the services and see what they need to," he said. "Maybe they need to see topology but not an event list. Or maybe they need to see latency but not necessarily packet loss. Our tool gives them that."Service-level agreements, those new contracting tools that help agencies gauge a vendor's promised performance, are also contributing to the rise of middleware products; they help service providers measure and report quickly key variables, such as network up time and latency.For federal agencies, with their heavily distributed infrastructures and numerous management platforms, middleware is a perfect fit, said Helen Donnelley Toth, vice president of marketing for Managed Objects, which recently began selling its product on the General Services Administration schedule through reseller NCI of McLean. "This type of tool really tries to level the playing field a little bit for in-house departments that usually have a shortage of qualified IT staff," she said. "This is a unified way to manage all your resources."For systems integrators, the new tools offer the same benefits as they do for in-house departments: the ability to do more with less staff; to monitor, troubleshoot and fix the network without always calling in a senior administrator, and to quickly measure network performance statistics.As a result, systems integrators who arm themselves with network management platforms and network management middleware products that are easy to use can expect to win additional business. In bidding situations, using tools can act as a discriminator, since they can provide value to an organization immediately. And, Ptak said: "They will allow systems integrators to build more robust solutions and more bulletproof solutions for their clients."Brown noted that, thanks to the complexity of networks, running them will never be so easy as eliminating the human factor from the equation, no matter how innovative the tools."But they provide the functionality we need to satisfy our customers," he said.XXXSPLITXXX-By Heather HayesEven as the private sector puts out new products to make it easier for in-house staff to manage networks, the public sector is, ironically, finally beginning to contemplate giving up control of their networks, according to industry officials. "We're seeing the whole nature of the business begin to move from a support-type structure to a service-type structure," said Jon Marcy, Navy Business Area Manager for Irving, Texas-based GTE Corp.'s Federal Network Systems.But while many agencies are thinking about the trend, most are taking a wait-and-see approach as the Navy wraps up the procurement for its massive Navy/Marine Corps Intranet project. That is a 450,000-seat, end-to-end solution that farms out responsibility for everything from the telecommunications links to local-area and wide-area network management to the software on the desktops."What it all means is somebody else gets to worry about upgrades and refresh rates, about what needs to be done to maintain the service and keep the infrastructure up to where the commercial market is going," said Joe Cipriano, program executive officer for the Navy/Marine Corps Intranet. "We want to let the experts do that for us in the most cost-effective way, and let us concentrate on what to do with the information and how to make better decisions faster with the information."Such thinking is a radical turnabout for most federal officials, who for years had sought to keep a tight hold on their networks, even as they began outsourcing their desktops through seat management contracts. Declining budgets and long-suffering difficulties with attracting and retaining top-quality staff are driving the new trend, but several other market factors are easing the transition to outsourcing and making it more palatable for federal agencies, according to Rich Ptak, vice president for systems and applications management for the Hurwitz Group, an IT research firm in Boston.For starters, organizations now have a better ability to secure their enterprises thanks to the maturing presence of virtual private networks. And secondly, a larger number of companies now can take over networks, meaning agencies can get the service at an increasingly competitive price. In addition, federal agencies now can better gauge vendor performance through comprehensive and detailed service-level agreements."From an enterprise perspective, all of this is really going to expand the number of options for how they provide IT services within the organization," Ptak said. "And I think it's going to put increasing pressure on those enterprises that do have internal IT operations to be more efficient and more attuned to the needs of the business."For service providers, not only will the trend open the door to more revenue opportunities, but it will put more pressure on them to "create bundled services where you not only provide the network and the underlying management, but you do the management, and maybe even the applications," Ptak said. "It will put pressure toward consolidating around more specialized services."Marcy noted that systems integrators looking to manage networks for federal agencies need to be extremely sensitive to security concerns and to be prepared to implement the kind of monitoring, reporting and troubleshooting tools that will enable good measurement capabilities to satisfy service-level agreement requirements."Outsourcing the networks seems pretty inevitable to me," Marcy said. "It's been our experience that just outsourcing the desktops is not meeting the expectations some of these agencies were looking for as far as accountability. "If you're looking for end-to-end performance guarantees on your enterprise, a service provider has to own, operate and be responsible for the enterprise end-to-end, not just the end points of an enterprise," he said, "because the performance of the networks affects everything from the applications to the enterprise itself."