Optical/IP Networks

VPLS: The Future of VPNs?

Don't think the current telecom lull means there's nothing going on. There's a massive shift underway from circuit- to packet-switched networks. And carriers looking for a way out of the downturn will eventually have to buy their way out by investing in technologies that enable them to offer revenue-rich data services.

Enter Virtual Private LAN Service (VPLS), or Transparent LAN Service. An emerging subset of technologies, VPLS has captured the collective imagination of the telecom industry as a preferred alternative for creating Virtual Private Networks (VPNs), which allow corporate users to secure pipes through public networks.

In fact, VPLS has generated so much interest that Light Reading's paid research service, Optical Oracle, has devoted an entire report to the topic -- "VPLS Variations," released to subscribers this month.

VPLS stands out from alternative technologies, such as IP VPNs, because it seamlessly integrates carrier and customer networks, works with carriers' existing infrastructures, is simple to provision, and offers massive scaleability, particularly in routed networks.

Using VPLS, for instance, customers can connect multiple sites over a provider-managed IP/MPLS network. All sites using VPLS appear to be on the same LAN, regardless of their location. More importantly, VPLS uses the ubiquitous Ethernet protocol as the customer handoff, simplifying the LAN/WAN boundary and allowing for rapid and flexible service provisioning.

Despite all this, the road to VPLS deployment is anything but straight and narrow. The standard has been in development for the last two years, but the technology is dependent on the rollout of Ethernet services. What's more, vendors are wrestling over the best methods of implementation.

Products, which are just starting to roll out, conform to one of two drafts: Lasserre-V.Kompella and Kompella. In a scenario typical of all great standardization efforts, there's lots of drama, as a cast of characters that includes two brothers ranged on opposing sides (see Kompella vs Kompella) debates issues of scaleability, device discovery, and signaling.

On one side are the majority of vendors, which are adopting the Lasserre-V.Kompella draft. This list includes: On the other side, implementing the Kompella draft (which many of the aforementioned vendors helped author) is Juniper Networks Inc. (Nasdaq: JNPR).

Going it alone is quite a gamble for Juniper, but the vendor was one of the first players to develop VPLS technology -- and once you head down a certain path, it’s hard to reverse course. Juniper argues that it's making provisions to back both technologies in its equipment.

Who will win? Who will lose? The stakes are clearly delineated in the 18-page Optical Oracle report, which describes in detail what VPLS is, how its two key drafts compare with each other and other technologies, and what differentiates the protocol. One section provides a detailed product assessment and financial analysis of leading vendors as a basis for handicapping the market.

"VPLS Variations" is available for sale here. A single-user subscription to the single report is $400. An annual subscription to the service, which costs $1,250, grants a user access to this report as well as the entire archive of past Optical Oracle Research -- and any new reports for the next 12 months.

— Mary Jander, Senior Editor, Light Reading

Editor's note: Light Reading is not affiliated with Oracle Corporation.

vach_kompella 12/5/2012 | 12:24:01 AM
re: VPLS: The Future of VPNs? Surprisingly, you didn't mention that TiMetra is doing VPLS. Hopefully, that was an omission.

Let's stay away from the drama and focus on technology, please? Thanks.

Mary Jander 12/5/2012 | 12:23:59 AM
re: VPLS: The Future of VPNs? Yes, I was taking a shorthand approach. TiMetra and some other firms are mentioned in the report. I've added them in to the article now.
signmeup 12/5/2012 | 12:23:51 AM
re: VPLS: The Future of VPNs? Considering the architect of one of the 2 VPLS drafts works for TiMetra, I would hardly state that it was a simple oversight. Your response?

BTW: I am not affiliated with any of the companies listed in the article; however I find it interesting that one can do an article on VPLS and not mention one of the technology leaders in that space.
alcabash 12/5/2012 | 12:23:48 AM
re: VPLS: The Future of VPNs? I was trying to find out about TiMetra's financials, how much did they raise? when was their last round ?
Are they for sale?
Scott Raynovich 12/5/2012 | 12:23:48 AM
re: VPLS: The Future of VPNs? >Considering the architect of one of the 2 VPLS >drafts works for TiMetra, I would hardly state >that it was a simple oversight. Your response?

It was. Timetra is included in detail in the Optical Oracle report about which this story is written. They are discussed at length.
cc_junk 12/5/2012 | 12:23:06 AM
re: VPLS: The Future of VPNs? VPLS indeed should be far easier to order and provision than BGP/MPLS VPNs. One just needs specify the access circuit and VPN, and possibly 802.1q classes. No IP addressing, protocol, BGP attributes, etc.

On the other hand, a VPLS VPN does not scale to a large number of sites. You would not put 50 or 100 much 1,000 sites onto a single LAN segment. It is not known if you can run, for example, that many OSPF peers on a single LAN subnet (will the designated router algorithm scale well with a large number of routers on the subnet?).

But it should fit very well the market for enterprises with a small number of sites - perhaps up to a few dozen?
tmc1 12/5/2012 | 12:23:04 AM
re: VPLS: The Future of VPNs? cc_junk,

H-VPLS will easily allow vpls to scale to 100s or 1000s of sites by reducing signalling and replication demands on one PE significantly. similar to route reflection in 2547 but less complicated.

Sign In