& cplSiteName &

MPLS Mavens Tout Progress

Light Reading
News Analysis
Light Reading
3/27/2001

This week marks the completion of two high-profile interoperability tests of equipment and software supporting multiprotocol label switching (MPLS), the peer-to-peer traffic exchange protocol established for routers by the Internet Engineering Task Force (IETF).

The technology, employed by many networking products under development, is expected to be adopted as a method of engineering faster traffic routes along high-speed networks.

On Monday, the University of New Hampshire InterOperability Laboratory announced that 11 vendors have completed MPLS tests there. And another round of tests is slated to wrap up this week at the Advanced Internet Laboratory at George Mason University.

The announcements highlight what should -- and shouldn't -- be expected of industry interoperability tests.

Unlike stress tests, such as the recent test of Internet core routers conducted by Light Reading (see Light Reading Testing Site ), the UNH and GMU tests are geared to help participants develop and sell their products, not gauge their performance.

Such tests function as development and sales tools for the vendors. Indeed, participants pay for the tests as an extension of their own R&D. (At UNH, members stump up $15,000 apiece, a bargain compared with the cost of privately commissioned tests.) Because corrections and adjustments are made to specific products during the tests, participants want the details kept secret.

"We use the tests to make sure our interpretations of the protocol are correct," says Ananda Sen Gupta, manager of MPLS products at Agilent Technologies Inc. (NYSE: A) and vice-chair of the interoperability group at the MPLS Forum.

UNH confirms this. "We varied the test bed quite a bit," says Rob Blais, MPLS Consortium manager at the UNH lab. In the first round, he says, the tests were based on passing data between specific devices. As that was ironed out, a larger network was built to which every participant was linked. In the end, "Things went pretty well. We saw a pretty impressive level of interoperability."

The tests also function as marketing tools. For example, Avici Systems Inc. (Nasdaq: AVCI; Frankfurt: BVC7) says it brought some of its customers to last year's MPLS tests at GMU (under NDA, of course) to demonstrate that its routers could work in a multivendor network.

An added benefit of testing at labs like UNH's and GMU's, sources say, is the creation of suites of interoperability and conformance test software that can be used to determine how well a product meets a spec -- in some cases for a fee, in others for free. But in the case of MPLS, sources say it will be awhile before any final packages are ready. "Nothing's finalized," says Blais.

The UNH testbed included routers from Alcatel SA (NYSE: ALA; Paris: CGEP:PA), Avici Systems, Cisco Systems Inc. (Nasdaq: CSCO), and Unisphere Networks Inc. (Nasdaq: UNSP); test gear from Agilent, GN Nettest, Network Associates Inc. (Nasdaq: NETA), and Spirent Communications; and protocol software for components and subsystems from Data Connection, Netplane Systems, and Virata Corp. (Nasdaq: VRTA).

The list of GMU test participants wasn't public. However, Avici and Nortel Networks Corp. (NYSE/Toronto: NT) confirmed their involvement. (Nortel said it didn't have time to do both GMU and UNH.) And Avici Systems said Cisco and Unisphere also participated. Juniper Networks Inc. (Nasdaq: JNPR) says it chose not to participate in either the UNH or GMU tests this year because it didn't feel like allocating the considerable resources (including several staff members) normally required to complete the tests.

The testing of MPLS raises the question of when and how interoperability tests will be conducted of Generalized MPLS (G-MPLS), the IETF's spec in progress that defines MPLS over a range of transports, including optical wavelengths. "We're investigating G-MPLS testing" says UNH's Blais. The trouble is, he asserts, that even though vendors want tests, they don't have any G-MPLS implementations to test.

Instead, it looks like the next round of optical interoperability tests will focus on the Optical Internetworking Forum (OIF). That group plans a demonstration of its User Network Interface (UNI) signaling at the Supercomm trade show in June. Preceding the demo, a closed-door interoperability test will take place at UNH. While Agilent, Avici, and Nortel say they'll be involved, the full details of who'll be included and what the scope of the tests will be aren't yet known.

-- Mary Jander, Senior Editor, Light Reading http://www.lightreading.com

(5)  | 
Comment  | 
Print  | 
Related Stories
Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
Scott Raynovich
Scott Raynovich
12/4/2012 | 8:39:04 PM
re: MPLS Mavens Tout Progress
strange reading of the article, I thought it was pretty straight and to the point.
Hanover_Fist
Hanover_Fist
12/4/2012 | 8:39:04 PM
re: MPLS Mavens Tout Progress
LightReading, the National Enquirer for the technology industry has done it again.

Only they could find fault with Interoperability testing - probably the most harmless and helpful events orgainized in the industry.

LightReading, for those light on intelligence who liket to waste time reading meaningless dribble.
Steve Saunders
Steve Saunders
12/4/2012 | 8:39:03 PM
re: MPLS Mavens Tout Progress
Hanover_Fist,

Back again? Can't say we missed you.

And this post is certainly up to your usual standard.

"LightReading, for those light on intelligence who liket to waste time reading meaningless dribble."

Interesting comment, given how much time you spend on the site. And I think you mean drivel, not dribble.

You no like? Leave.

What's up with that user name, anyway?

Steve
lagreuse
lagreuse
12/4/2012 | 8:38:40 PM
re: MPLS Mavens Tout Progress
Hanover_Fist, are you feeling OK? Your message (not to mention your pen name) is one of the most puzzling I have seen on lightreading for some time.
Dutch
Dutch
12/4/2012 | 8:34:34 PM
re: MPLS Mavens Tout Progress
Not exactly the comments youGre looking for butGǪ

Over the years, IGve acquired a lot of useless knowledge. I know what the number Gǣ33Gǥ on a bottle of Rolling Rock means, who Kilroy was (and where he was), and the average logarithmic energy decrement per collision of water is, given a fast neutron, among other things. I believe Hanover Fist is a character in an 80Gs era cartoon flick. If I remember correctly (itGs been 15 years) heGs a timid creature, who goes ballistic upon questioning. (There may be some hidden meaning in that, but IGve no desire to pursue it.) The movie has a great soundtrack (Blue Oyster Cult, Sammy Hagar, etc.), some solid violence, and a bit of skin, but the plot is a little weak. Great stuff when youGre 14, (or half lit).

Great site.

Dutch
Featured Video
Upcoming Live Events
December 3-5, 2019, Vienna, Austria
December 3, 2019, New York, New York
March 16-18, 2020, Embassy Suites, Denver, Colorado
May 18-20, 2020, Irving Convention Center, Dallas, TX
All Upcoming Live Events