SAN FRANCISCO - Planning for a crucial virtual private network (VPN) installation at a major corporation is not a pretty process.
You spend a lot of time talking with the workers in your business trying to determine their technology needs and business goals, and even more time trying to get vendors to clearly explain what their products do. After all that, you might decide to pass on the technology.
Those were two of the lessons learned by Detroit automobile manufacturer General Motors Corp., which has initiated a VPN project that will eventually link tens of thousands of users, and by CIGNA Corp., a health insurance and financial services company in Philadelphia.
Norm Samuels, chief security architect at General Motors, and Len Dorrian, vice president of telecommunications TAP at CIGNA Systems Division, dispensed their advice and shared their insights as customers into the state of the virtual private network market at the Catalyst Conference 2002. The annual networking event is hosted by the Burton Group, networking consultants and researchers in Midvale, Utah.
Samuels said like all large companies today, General Motors not only faces the challenge of opening up its networks to communicate with partners and employees, but also keeping them as secure as possible. He decided a virtual private network was the simplest and cheapest way to go.
"General Motors is not the CIA and GM is not going to solve world hunger so GM chose the best alternative," he said.
His first step was to learn the requirements of the people who would use it, such as the remote workers who General Motors wanted to tie into the network.
"What the road warriors used to do is come home, start downloading data, eat some supper, go to bed and wake up in the morning hoping the downloading was complete," he said.
CIGNA's Dorrion said his company started looking at an Internet Protocol VPN in early 2001 and has yet to choose a vendor.
His company set up a team of 15 to 20 people to learn about the technology and seek advice from analysts, equipment vendors and carriers. The team whittled down vendors to a group of four and then looked at how their products would work and what value each would bring to the business, such as cost savings and manageability.
Dorrion said the process turned up some troubling trends. It became clear that often the marketing people trying to sell the products didn't understand the technology, making it difficult for his evaluation team to get its questions answered.
Also, he said, sometimes product managers from the same vendor would wind up competing against each other for CIGNA's business, which added to the confusion. Dorrion said that vendors did not appear to have much experience with the converged network that CIGNA was considering.
"They all wanted to give you the dazzle diagrams" he said, but the technology answers just seemed to fall short.
"CIGNA does not want to be leading edge," he said.
Citing WorldCom's accounting and financial troubles of the past weeks, he said the final choice of a vendor will depend on how WorldCom and AT&T are performing when CIGNA decides to go ahead.