This week I attended and presented at the Future-Net conference in Boston. This conference isn't one of the biggest conferences but this intimate event provides some great presentations and the opportunity to meet and share ideas with a lot of great people. I attended this conference last year and provided a review of what I saw. I wanted to share with you some of what I learned this year and discuss the highlights of the conference.
Future-Net is a conference that has been taking place for many years. It started out as the MPLScon and focused on networking infrastructure operated by service providers. Last year's event focused on standards being created by the IP/MPLS Forum and the Metro Ethernet Forum (MEF). FutureNet is now the primary conference for the Broadband Forum. As the technology landscape changed over recent years the Future-Net conference has continued to remain relevant and focus on Metro Ethernet, layer-2 VPN technologies like VPLS, T-MPLS, PBT. In recent years the conference has also started to cover green technologies, Session Initiation Protocol (SIP), LISP (more on this later), and IP Version 6 (IPv6).
There were several presentations on IPv6 and its inevitability. John Curran (President and CEO of the American Registry for Internet Numbers (ARIN)) talked about the rapidly approaching end of available IPv4 addresses. His presentation covered what service providers, enterprises, content providers, and manufacturers should be doing to make sure they are focusing on the transition. The panel that I was part fielded questions from the audience on IPv6 and how to migrate to it. The panel was moderated by Johna Till Johnson and comprised of myself, John Curran, Doug Junkins (Vice President of IP Development for NTT America's Global IP Network), John Jason Brzozowski (Architect and Principal Engineer at Comcast), and Fernando Gont.
I spent most of my time at future-Net in the Internet track which covered IPv6, Metro Ethernet, MPLS-TP, VPLS, PBB/PBT, and service-provider solutions. There was another parallel track that covered enterprise topics like SIP and green computing/sensors/smartgrid. I will have to go back and review these presentations because they are very important topics. I must admit, it was difficult to choose at times which track would be best to attend because they were both filled with great information and ideas.
One of the topics that was covered, by Darrel Lewis of Cisco Systems, during the first day's tutorials was Locator/IS Separation Protocol (LISP). LISP is a new strategy so split the IP address into two pieces; one identifies the device (Endpoint Identifier (EID) and one identifies the attachment point (Routing Locator (RLOC)). LISP has been an active working group within the IETF for a number of years now. The latest draft is draft-ietf-lisp-07.txt. The benefits of LISP is that it allows the current single IP address field to improve routing table scalability for multi-homed organizations and will provide for any-to-any connectivity. Cisco Systems is supporting LISP and has already worked on router code for the protocol. LISP will be available to their customers in IOS release 15.1.
One of the highlights of the conference is the "beer-and-pizza shootout" which took place Tuesday night. This is a panel that typically covers some controversial topic facing our industry. This year Johna Till Johnson (President and Senior Founding Partner Nemertes Research) provided the topic of conversation "Is the Internet broke?" The topic was basically to stimulate discussion about which new protocol can overcome the deficiencies of the current IPv4 Internet. The three competing protocols are IPv6, LISP, and PNA. PNA is an acronym for a book written by John Day titled Patterns in Network Architecture. The book covers how networking protocols should be constructed and use addressing schemes that provide hierarchy and scalability. The people working on these alternative-Internet protocols have formed The Pouzin Society. These ideas have lead to discussions of alternatives to IPv4 and IPv6 addressing and to the development of Recursive Inter-Network Architecture (RINA).
The central theme of the beer-and-pizza shootout was to discuss what alternatives to IPv6 exist to help fix the addressing and Internet routing table scalability. It was a lengthy and healthy debate but the conclusion seemed to be that in the short term the world has no alternative but to implement IPv6. The growth of the IPv4 routing table and the fact that IPv6 does not fix the multihoming problem are cause for concern about the future. However, in the meantime we should put effort into determine what the future of the Internet should look like and how to get more scalability out of the current IPv4 Internet.
I had a great time attending and being part of the Future-Net conference this year. I learned a lot from the industry-experts presenting the latest information on network technologies. I am looking forward to next year's conference. There is no doubt that IPv6 and LISP will become increasingly important topics.