Work remains, but the path to a single network is getting clearer.Exempla Healthcare’s three-hospital network uses an OC-3 ring to link about 15 buildings in the Denver area. “We run 225 different applications,” says Exempla CTO Lots Pook. “And 45 or so have to run on a stand-alone network because they can’t ride our IP network.”Even networks within buildings have to be doubled up to meet legacy demands. For example, fetal monitors have to be wired separately using Category 5 cable because they run a 16-volt pulse signal. In the legacy-heavy, highly regulated healthcare industry, such applications are not uncommon.Also: Ups and downs – the benefits and drawbacks of IP integration In its newest facility, Pook’s team is getting a taste of the good life, with merged voice and data and a higher percentage of applications running on IP. “I can run that at a much lower cost,” Pook says. However, Exempla won’t be all-IP any time soon, as long as healthcare companies require specialized networks to meet patient and regulatory needs.While the industry at large marches toward an integrated IP future, most organizations still have to deal with a collection of legacy networks. How many? Even when you don’t have to support special equipment, there can be as many as a half-dozen, according to Nick Lippis, CEO of consultancy Lippis Enterprises. In addition to voice and IP, there’s still some semblance of SNA networks in most companies, say nothing of video networks, dial-up networks, one devoted to alarms and physical security, and a wireless network (albeit one that’s typically leased).While the assumption is that, over time, all traffic is collapsed onto IP, questions remain about everything from the motivations and risks to the impediments to integration. The migration is slow, steady and methodical.Gartner analyst Bob Hafner says there’s a slew of activity around IP migration, but nothing resembling a tidal wave. “Has anybody moved all their networks to a single network? Not yet. We’re in the early stages. Key opportunities are being addressed, but you don’t see an overall strategic plan.”The key opportunity of choice, not surprisingly, is IP telephony. In informal surveys, Gartner finds 80% of companies have some sort of IP telephony program in place, “but it may be anything from a trial with five phones to an entire 2,000-person location,” Hafner says.Contributing factorsAmong corporations migrating to IP, cost is the primary driver. In 2003 when the Cancer Therapy and Research Center installed storage-area networks in its buildings in San Antonio, Texas, CTO Mike Luter opted to link them using iSCSI rather than the more traditional Fibre Channel, meaning the storage traffic could ride the IP network. Cost was the biggest factor, Luter says. The center’s buildings are 22 miles apart, and “to do Fibre Channel, we would have had to put another fiber in. With iSCSI we were able to use existing fiber.”The center installed Cisco SN 5428 storage routers in both locations to consolidate storage resources on shared EMC Clariion arrays, and the 5428s are linked via iSCSI over the IP network.Because the Cancer Therapy and Research Center is a not-for-profit organization, its purchase of the Cisco gear was closely scrutinized. The reduced cost of ongoing maintenance for a single-protocol network was a major factor in the approval, Luter says. This echoes the sentiments of nearly every IT group considering IP integration: The major appeal is to create one infrastructure, reducing the need for different skill sets.Some of the cost benefits of IP integration – such as a significant drop in the price of telephone moves, adds and changes when companies adopt VoIP – are well documented. Others might take longer to pan out, but could be more important in the long run. For example, Exempla’s Pook says the company’s all-IP building frees up a full-time network technician’s salary. The primary reason for IP integration is savings, Hafner says. “We haven’t seen any scenarios in which enterprises gained competitive advantage. However, we have seen cases in which it became much easier [for companies] to do things they wanted to do because of integration,” he says.In fact, sometimes the migration to IP is the only feasible way to run needed applications. That’s the case at the School District of Lancaster, Pa., where network manager Tim Fake is leading a shift from an infrastructure that can be generously described as piecemeal to pure IP.With more than 11,000 students, the district has 21 buildings linked via a 100M bit/sec ATM network and a separate voice network. Fake’s group is nearly finished with the first phase of a shift to an all-IP infrastructure that will be built around a Cisco Catalyst 6500 in the data center; a Cisco Catalyst 4500 Series for each campus; and Catalyst 2950/3550s in wiring closets.While Fake welcomes the simplified infrastructure, the big driver for the school district is the ability to offer “apps we just cannot do right now that are really crucial for education,” he says, including video on demand and distance learning. Ultimately, Fake says, he will be able to offer VoIP, video on demand, 100M bit/sec Internet access to every desktop and unified security.In addition to cost and new applications, analysts say senior business management is aiding the migration to IP by calling for things that are enabled by it. “CEOs and CFOs may not understand the network, but they understand the importance of the network,” Lippis says.They see that profit drivers such as improved collaboration, easier access to enterprise applications and better customer support can be “goosed” by network technology, Lippis says. “This drives the push toward IP because it affects how a company interacts with its entire ecosystem” of suppliers, partners, employees and customers, he says. “It’s easy to see how a single infrastructure makes this doable.”RisksRegardless of the potential of IP integration, IT organizations want assurance about security and reliability. Is one IP infrastructure even a desirable thing given single point of failure concerns?Gartner’s Hafner says not to worry. Proper analysis of network traffic, followed by appropriate network design, can address such issues.Kim Perdikou, CTO at Juniper, agrees. “The more traffic there is [running over IP networks], the more analysis is done on the business risk,” Perdikou says.Companies have found a variety of ways to ensure uptime in their integrated IP networks. The Cisco storage routers at the Cancer Therapy and Research Center include a feature that lets the center boot servers directly across the IP network.According to Luter, that not only improves the organization’s patient service, but lets it stick to a demanding schedule that has complex medical procedures occurring every 10 minutes. “We need to get back up as quickly as possible,” he says. “With Network Boot, we can be up and running after a server failure in 10 to 20 minutes.”That’s a vast improvement over the two to eight hours it would take to manually install an operating system and download up-to-date data onto a new server with tape backup.Because everyone’s network and business drivers are different, there is no central migration plan that companies can gauge their IP integration efforts against. But analysts, users and vendors agree that companies with no IP integration plan whatsoever are lagging behind.For companies looking for new ways to streamline costs, management and gird for the future, an aggressive move toward IP integration might be an option to consider.Pure integration probably won’t happen anytime soon, but the tectonic plates have shifted and the benefits of integration outweigh the drawbacks, so the march toward integration is better undertaken sooner rather than later.Ulfelder is a technology and automotive writer in Southborough, Mass. He can be reached at sulfelder@charter.net. Related content news Mainframe modernization gets a boost from Kyndryl, AWS collaboration Kyndryl and AWS have expanded their partnership to help enterprise customers simplify and accelerate their mainframe modernization initiatives. By Michael Cooney Nov 30, 2023 4 mins Mainframes Mainframes Mainframes news AWS and Nvidia partner on Project Ceiba, a GPU-powered AI supercomputer The companies are extending their AI partnership, and one key initiative is a supercomputer that will be integrated with AWS services and used by Nvidia’s own R&D teams. By Andy Patrizio Nov 30, 2023 3 mins CPUs and Processors Generative AI Supercomputers news VMware stung by defections and layoffs after Broadcom close Layoffs and executive departures are expected after an acquisition, but there's also concern about VMware customer retention. By Andy Patrizio Nov 30, 2023 3 mins Virtualization Data Center Industry news US will take decades for supply chain independence in chips: Nvidia CEO Jensen Huang pointed out that Nvidia’s latest AI servers have 35,000 parts from all over the world, including Taiwan. By Sam Reynolds Nov 30, 2023 4 mins CPUs and Processors Technology Industry Podcasts Videos Resources Events NEWSLETTERS Newsletter Promo Module Test Description for newsletter promo module. Please enter a valid email address Subscribe