It’s not particularly new, and a lot of people have no idea what it is. Nonetheless, it is still one of the most successful of IETF’s protocol suites. Most of the big Internet applications that are actually making money are on the Multi-Protocol Label Switching (MPLS) protocol. And of course, when there is money involved, people will eventually start to pay attention.
MPLS evolved out of asynchronous transfer mode (ATM) in the ’90s, when the IETF integrated ATM’s convenient label-switching capability, with IP. ATM of course, was immensely useful but costly to implement, and marrying it with the Internet Protocol was a stroke of genius. Besides its great flexibility, MPLS is also protocol-neutral; in other words, it works, as its name would indicate, in a multi-protocol environment. It is often used in larger installations such as metro Ethernet or in high-performance telecom networks. The second part of the moniker, the “label switching” part, simply means that the individual data packets are transferred to their ultimate destinations based on the contents of a label, rather than the entire packet—which is more efficient and reduces latency.
Now that we have a good idea of what MPLS really is, the next question is, why do we really need it? A recent blog entry on INE answered that question. INE Instructor refers back to that IETF working group, which first met in 1997, where the engineers identified four main areas they wanted to address with the new protocol. Those four main goals are still relevant today, and they are: Scalability of network layer routing, more flexibility, better performance, and easier integration of routers with cell-switched technologies.
I would suggest that a fifth unstated goal has also been met, and that is: Do more things with less money. Engineers at IETF working groups are usually more concerned with the details of technology than they are with commercial applications and saving corporations large amounts of cash, but their pure motives notwithstanding, that is in fact the overriding impetus behind it. And without a commercial reason for its existence and a solid ROI, it never would have taken off.
And this may be a tangential argument, but I remember in the early days of the Internet, the first users were purists who saw the Internet as a strictly non-commercial tool to be used for academic purposes and sharing ideas. A noble idea indeed. Those pre-dotcom purists didn’t foresee, and didn’t want, advertising, commercial web sites, glitzy Internet apps and Internet marketing. But the cold hard facts are this: Were it not for those commercial applications, the Internet simply would not have evolved to the level of sophistication we see today, and we would still be using a text-based browser over dial-up connections. The greatest goal of the Internet now is as it should be: Allow more people and more companies, to do more things, more efficiently and with less money. The development of MPLS fits in nicely with that sentiment.
RELATED
The Falling Costs of Wide Area Networks
MPLS Pricing
Request MPLS pricing from 3 providers
I totally agree. The commercial application of MPLS has been the major driver of its success.
While technically there are many reasons to use MPLS within a Service Provider’s network the ROI comes from the huge product potential that MPLS brings to a Service Provider’s portfolio, allowing Service Providers to offer Layer 3 VPN, and more recently Layer 2 VPN (VLL & VPLS) services.