A look at some of networking’s most enduring – and contentious – power struggles.
Some things never change. Year after year, some of the same network vendors lock horns time and again, trading off successes and failures in a never-ending cycle of power one-upmanship. Or service providers battle their own business and technology demons ceaselessly. We look at the latest developments in four such perennial fights and share how they will affect your business over the next year and beyond.
Cisco vs. the world
For nearly 10 years, we have extolled Cisco’s market dominance and praised CEO John Chambers’ savvy while pondering the challengers eyeing the company’s networking crown. Although many have made a decent run at Cisco, none to date have made more than passing progress.
This time around, Juniper has taken a few pages from Cisco’s own playbook to launch a formidable offensive. Not content simply to compete with Cisco in the high-powered carrier-class routing market, Juniper in February began an all-out assault on Cisco’s other core market, the Fortune 200. In its first salvo, Juniper bought enterprise-focused security and VPN vendor NetScreen Technologies.
“NetScreen gives Juniper a really strong leg in the enterprise. In particular, it can now play in the security and collaboration spaces, which are areas that are hotly growing and are next-generation, application-focused solutions,” says Johna Till Johnson, president of Nemertes Research and a Network World columnist.
Juniper fired again at Cisco in the spring with the launch of the Infranet Initiative. Infranets promise to revolutionize the market for public IP services by attacking, in a standard way, the Internet’s greatest faults: security and reliability. The Infranet Initiative has received much support from several industry segments. Members include global service providers such as AOL, British Telecom, Deutsche Telekom, France Telecom, Level 3 Communications and Qwest; network gear vendors Ericsson, Lucent and Polycom; and application and computing companies such as HP, IBM and Oracle.
Cisco has yet to jump on the infranet bandwagon. Maybe it doesn’t want a new initiative to dull the luster of its own technologies. Or perhaps it would rather not join a group it isn’t leading. Either way, Cisco faces a choice: adapt to the new competitive environment or lose market dominance.
“Cisco is definitely at risk to Juniper’s ability to promote infranets as a revolution,” says Thomas Nolle, president of consultancy CIMI. “I’d say Cisco now is where IBM was in the 1980s. It’s a giant in a field that isn’t getting bigger fast enough. And like IBM, Cisco will have to adapt and remake itself in order to remain successful.”
No one’s counting Cisco out just yet. “I don’t see Cisco losing its market edge,” says Mark Bieberich, program manager at The Yankee Group. “Right now, Cisco holds 70% of the total market. I would be very surprised to see its share decline very much, if at all, in 2005. It’s just not going to happen.”
Johnson concurs. “You can never rule Cisco out. It is in a bit of a lull at the moment, but you never know when it will pick up again.”
The upshot: Juniper has the edge now, but stay tuned.
Long-distance carriers struggle for survival
The merger between Sprint and Nextel announced Dec. 15 underscores how heated the struggle for survival has become among long-distance carriers. With razor-thin margins and new technologies such as wireless and broadband siphoning off what used to be considered safe revenue streams, today’s remaining carriers are battling for their lives. And 2005 promises to be the year when we’ll see whether the final two, AT&T and MCI, enter into their own acquisitions or, perhaps, wither away on the vine.
The newly named Sprint Nextel will be a formidable competitor. With combined revenue of about $40 billion, the company will be a powerhouse in wireless, with more than 35 million subscribers. It also will spin off its local telecom business, merger documents say.
AT&T is an iffier proposition. With its well-publicized exit from the consumer business and its decision to lay off 7,500 employees and write down assets by $11.4 billion in the third quarter of this year, the situation looks shaky. But on the whole, those moves open up more alternatives and might just make the business salvageable. Such moves nudge the door open for potential buyers by making AT&T a bit smaller and more palatable. Until now, few have believed AT&T would make a good merger candidate because any offer would trigger regulatory delays and costs.
But beyond that, AT&T is now free to focus more readily on its core business – serving large corporations. New initiatives like its Concept of One, which promises to unify network infrastructures by consolidating legacy ATM and frame relay networks over a Multi-protocol Label Switching backbone, are key to AT&T’s survival, as is good management. And AT&T seems to be lucky in both regards.
“Top-line revenue numbers may shrink a little because it’s losing some consumer business, but at the end of the day, it’ll be running a more profitable business,” says David Parks, senior analyst at Yankee. “It needs to focus on next-generation capabilities like Concept of One and on trying to differentiate itself, and I think it is.”
Johnson agrees, adding that AT&T CEO David Dorman has what it takes to keep the company viable. “Part of AT&T’s problem is that in the past, it has had people with limited vision and limited ability to execute, whereas Dorman is really good at both,” she says. “I’m pretty bullish on AT&T’s chances.”
Unfortunately, the same can’t be said for MCI. Having emerged from Chapter 11, conventional wisdom pegs the carrier for acquisition or death sometime in 2005. A possible buyer would be BellSouth, which is the only RBOC left without a presence outside its region. “With its debt paid down, MCI is a fairly attractive merger candidate, and BellSouth has been looking at it pretty hard,” Nolle says. “And my information is that the Department of Justice would be favorably disposed toward a merger between MCI and BellSouth.” He says talk of such a merger will circulate sometime in 2005.
Others say that would be crazy, because MCI would cost far more cash and trouble than it’s worth. “MCI’s network was cobbled together with something like 27 discrete networks, and it still has no integration, no operational efficiencies, and it can’t get a combined bill out to save its soul,” Johnson says. “Any self-respecting Bell will take a look at this and realize that with that much money, they could just go build it themselves.”
And that leaves just one alternative. “I see MCI withering away,” Johnson says. “At some point, some private equity firm or some other player will pick it up at a rock-bottom rate and put it out of its misery.”
The upshot: AT&T has an edge on survival over MCI.
Ethernet vs. traditional WAN services
ATM and frame relay are constantly pitted against “the next big thing” in the WAN, and for years they have pretty much vanquished all comers. But Ethernet advocates continue their efforts, most recently with Virtual Private LAN Service (VPLS). By providing an Ethernet interface for corporate routers and switches, VPLS simulates a LAN routing environment among corporate sites.
Experts say the smart bets are still on ATM and frame relay, at least for the next two years or so. Why? VPLS is simply not in the carriers’ best interests right now.
“You haven’t seen AT&T, MCI or Sprint roll out VPLS because they’re worried about it cannibalizing their tremendous installed base of revenues,” Yankee’s Parks says. “They do billions of dollars a year in frame and ATM, so they’d rather preserve what they can and migrate people to Layer 3 IP VPNs” vs. offering wholesale moves to VPLS, he says. That’s why only two providers – Masergy Communications and Time Warner Telecom – offer VPLS today. Neither has an installed base to protect. User organizations also may have little incentive to move unless the carriers can price VPLS ultra-attractively. And that’s not likely to happen.
“Enterprises have told me they’ll replace frame relay and ATM with anything that’s 30% to 40% cheaper,” Nolle says. “For the service providers, there’s no incentive to offer that kind of pricing.”
Still, some may find VPLS’ features, especially its bandwidth, too attractive to pass up. “You could get a 10M bit/sec Ethernet VPLS service at a location for about what a frame T-1 costs you. Plus, you get any-to-any connectivity,” Parks says. VPLS is a Layer 2 technology, which might be a selling point for organizations that want to retain control over their own routing and are hesitant to move to a Layer 3 setup.
On the downside, VPLS requires an Ethernet connection at every endpoint. This means that VPLS could require changing out the premises router, which would make migrating to an IP VPN more attractive, because it could leverage the old router and the old frame or ATM port.
The upshot: VPLS eventually might make Ethernet a winner in the WAN, but in the next year or two, the edge is with ATM and frame relay.
Retention vs. storage management
Now that deadlines for various compliance regulations, such as the Sarbanes-Oxley Act and Health Insurance Portability and Accountability Act, are here, IT executives face technology decisions that, if not handled properly, could land top corporate managers in jail.
However, by all accounts, most organizations are still struggling against the tide here. They fall into two camps: those that understand the issue fully and are working doggedly to deploy storage mechanisms and policies to comply with the regulations, and those that believe they are compliant simply because their policies meet the strict letter of the law. At this writing, both groups seem to be losing the battle.
“There’s what the law says, there’s what your policy says, and there’s what you really have to do to survive in the real world,” says Johnson, adding that most large organizations have yet to realize the difference.
An organization might have clear retention policies that state that after three years it throws everything out. That might adhere to the letter of the law, but what happens when that company is slapped with a suit and asked to produce documents older than the three-year limit?
Retention policies that dictate which documents are saved and which are deleted are “not going to protect you,” Johnson says. “In fact, when you get up in court and say you threw [a document] out because ‘we have a policy for throwing it out,’ the jury will convict you,” she says, citing the experience of one of her clients.
Does that mean companies need to save everything forever? Well, no. But they need to realize the scope of their retention policies and address their liabilities wisely. “You need a policy that says when you throw something out, how you throw it out and what you do if you ended up throwing it out incompletely,” she explains. “If your policy is to throw out all e-mails, how do you guarantee that e-mails haven’t been stored locally?”
Organizations that do understand the ramifications quickly become overwhelmed by the magnitude of the problem. Some hope that a new breed of compliance software tools is the answer.
“To my knowledge, there is no software anywhere that can guarantee you’ll be compliant,” Johnson says. “None, period.”
Because the regulations and the technologies they address are in a constant state of flux, investing in people is wiser, she says. “You can spend a lot of time and energy buying software, figuring it out and configuring it. But whether consultants or in-house staffers, the reality is that nobody knows how things like [Sarbanes-Oxley] will be interpreted. You really need to have human intelligence guiding this thing right now.”
The upshot: Dead even.
Cummings is a freelancer in North Andover, Mass. She can be reached at jocummings@comcast.net .
