x
Optical/IP

Optical Ethernet Angst

SAN JOSE, Calif. -- Ethernet's progress in carrier networks has taken center stage more than once. But at the Opticon 2001 conference held here this week, experts didn't seem to have the answers to questions that potential customers might have.

Presenters here waxed eloquent on the benefits of Ethernet in metropolitan area networks -- including increased capacity and control for carriers, along with cost savings from reduced operational expenses. But few have been able to say just how and when Ethernet will start to dominate the metro scene.

Take the specifications, for instance. Talks yesterday afternoon stressed the importance of MPLS (multiprotocol label switching), Resilient Packet Ring (RPR) protocols, and 10-Gbit/s specs in development at the Institute of Electrical and Electronics Engineers Inc. (IEEE). But all of these specs are still being worked on and are potential political minefields.

Approval of the IEEE specs, for instance, has been pushed back at least a couple of months. Originally slated for ballot late in 2001, the specs won't make it to final approval until March 2002, according to Jonathan Thatcher, principal engineer at World Wide Packets Inc. and chair of IEEE's 10-Gig Ethernet committee (802.3ae).

This doesn't mean anything is amiss, he asserts: "The spec's in great shape, terrific shape." Getting the final details nailed down will take a bit longer, however, than first anticipated.

Thatcher spoke on Optical Ethernet during an afternoon session yesterday, along with co-presenter Bruce Tolley, manager of emerging technologies at Cisco Systems Inc. (Nasdaq: CSCO) and VP of the 10 Gigabit Ethernet Alliance (10GEA).

Neither panelist could answer queries about just when 10-Gbit/s Ethernet would start to emerge in full force on carrier networks. "It will happen," Thatcher said. But many carriers, he noted, aren't willing to talk about their plans, since they're strategic to future success.

Some members of the audience seemed dubious. "We just spent a fortune to rebuild our network to compete with ILECs and IXCs," wrote one attendee on a card delivered to the podium. "Now you're saying we have to do this all over again?"

Later in the day, similar questions arose around an MPLS tutorial given by Chris Gibbings, an IP engineer with Level 3 Communications Inc. (Nasdaq: LVLT), one of the carriers most often cited as committed to Ethernet in the WAN. Gibbings told the audience that Level 3 was committed to starting Ethernet service throughout its network, possibly early next year. But he stopped short of specifics.

Despite the dearth of hard information on Ethernet's progress in carrier nets, many attendees and exhibitors seem convinced, like Thatcher, that it's set to rule the WAN. "We won't replace Sonet, but we think a gradual decline will begin as Ethernet deployments expand in 2003 and 2004," says Don Gibbs, CEO of VIPswitch Inc., which is set to start lab trials of its metro switch/router in September.



But when pressed, a few skeptics emerge.

"I'm not sure where the cost savings will be," admits Paul Liesenberg, VP of strategic marketing at ZettaCom Inc., a network processor company offering an OC192 traffic manager and switch fabric. "The optics are expensive for ten-gig, identical for Ethernet or Sonet," he notes.

Even if the cost of optics comes down, Liesenberg says higher utilization on networks will require those cards to be outfitted with more intelligence, making them more costly to produce.

Still, like others here, he seems willing to be shown. "We do think Ethernet will take over the access network," he maintains, even though Sonet will continue to prevail for a long time at the edge and core of the network.

— Mary Jander, Senior Editor, Light Reading
http://www.lightreading.com
jmd 12/4/2012 | 7:57:29 PM
re: Optical Ethernet Angst Times like these require more than good technology G a compelling business case is needed GǪ or you wait till the market thaws GǪ or bandwidth demand takes off legitimately.

The longer the market thaw takes the more startups will die. Regardless, the only way that more than a very few Metro Optical startups will survive is if the bandwidth demand (however itGs measured) comes close to the giddy projections made during the bubble era. Recent reports on the growth of consumer broadband access are encouraging but not nearly enough.
jshuler 12/4/2012 | 7:57:02 PM
re: Optical Ethernet Angst FYI, there will be at least one working draft proposal for RPR delivered at the IEEE 802.17 meeting in early September, representing the work of multiple companies. We are on track.
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE