In 2017, I realized a lot of the people I've worked with over the years were retiring. When these people leave the networking community, they take a wealth of knowledge about the intent, challenges, and inventions of the early Internet. I decided to capture as much of this history in oral format as possible--hence the history of networking recordings were started. I thought, at first, this would be a small, short-lived series, but I have been amazed by the reaction of the community, and the number of technologies and organizations involved in the design and operation of computer networks.
These recordings are released under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0). This means recordings can be distributed for any noncommercial purposes by anyone, so long as they are released in full (with no modifications).
Tony Li: BGP
Before BGP, the Internet core (or default free zone) went through several protocols. In this recording, Tony Li discusses the origin of the Border Gateway Protocol.
Daniel Walton: BGP Churn and Addpaths
BGP is a nondeterminstic protocol, which means it can converge on different paths depending on the order in which they are received. In this recording, Daniel Walton talks about the discovery and research around BGP churn, and the solutions the routing community adopted to solve the problem.
Daniel Walton: BGP Optimizations
BGP implementations did not always have the ability to scale to millions of routes; Daniel Walton joins the history of networking to talk through the history of optimizing BGP for scale and speed.
John Fraizer: BGP Route Servers and IXPs
BGP route servers play a critical role in the operation of an Internet Exchange Point (IXP). John Fraizer joins the history of networking to discuss how he developed one of the first route servers while building an IX in Ohio.
Geoff Huston: BGP Security
BGP, the protocol that "runs the Internet," is notoriously open to various kinds of attacks. Geoff Huston joins the history of networking to explain how we got to this point, and some of the attempts made at fixing it.
Terry Slattery and Bruce Pinsky: The Cisco Certified Internetworking Expert
Certifications in networking really started with the CCIE.
Terry Slattery and Rob Widmer: The Cisco CLI
Many different methods of configuration were available in the early days of configuring routers; Bay networks, for instance, provided on-screen navigation of SNMP MIBs. The CLI was a great improvement.
Dave Oran: DECnet
The story of DECnet labs is, in part, the story of how the inventions of some of the brightest minds in the early days of networking ultimately made their way into the Internet Protocol.
Mike Sullenberger: DMVPN
We often seem to think that new ideas, like SD-WAN, "just come out of nowhere." In this recording of the history of networking, Mike Sullenberger joins us to talk about DMVPN, which is one of the technologies that drove the SD-WAN revolution.
Rahul Aggarwal: eVPN
eVPN uses BGP to carry information to form layer 2 and layer 3 overlays. eVPN has expanded, over time, to solve most overlay problems.
Ross Callon: MPLS
Multiprotocol Label Switching, or MPLS, underlies much of the magic that makes large-scale transit networks run, and is a precursor to concepts like controller-based SDN.
George Swallow: MPLS/TE
Traffic engineering is an interesting case study in using packet based mechanisms to solve problems traditionally solved with circuit switched network technologies. George Swallow is one of the original designers of MPLS/TE.
Luca Martini: Psuedowires
In the first few decades of networking, circuit switched solutions reigned supreme. When packet switched networks were fully adopted, network engineers discovered there were some problems that were still best solved with circuit switched technologies--hence the invention of emulated point-to-point circuits over packets switched networks.
Jeff Tantsura: Segment Routing
Segment routing, according to many routing geeks, is what MPLS should have been in the first place. In this recording, Jeff Tantsura discusses the original use cases and design logic behind SR.
Stan Hanks: GRE
Paul Mockapetris: DNS
DNS is a critical system in all networks, as it converts names to topological addresses where services can be found.
Paul Vixie: DNS Adoption
Even though DNS seems obvious now, it took a good coder to sit down and write an initial implementation to get the protocol deployed in early networks. Paul Vixie, who wrote the original BIND toolset, joins the history of networking.
Mark Kosters: WHOIS
WHOIS is a crucial protocol for the day-to-day operation of the Internet, providing information on who owns what address space, and how to contact them.
Ken Duda: Backplanes
Many different kinds of backplanes have been used to connect line cards and other components in network devices like routers, switches, and firewalls. The original custom-designed busses gave way to plain Ethernet busses, then to crossbars and Clos fabrics. Along the way more exotic designs, such as virtual meshes on top of toroids have been used, as well. Join Ken Duda, Russ White, and Donald Sharp as we discuss the many different kinds of backplanes and the history of their development.
Yuval Bachar: Backpack
The concept of disaggregation could not be taken seriously before the rise of merchant silicon. Yuval Bachar designed the Backback, one of the first "white box" routers in the world.
Tony Li: Silicon Switching Engine
Software switching on general purpose processors was the rule in early networks; in fact, one of the original differences between a router and a switch was whether packets were switched in hardware or software. In this recording, Tony Li discusses the first hardware switching router, the SSE, which he designed.
Dinesh Dutt: Divergence of Routing and Compute
In many ways, routers and switches are just special purpose computing platforms. How did they come to be considered completely different kinds of devices? Dinesh Dutt joins the history of networking to explain.
Kevin Herbert: Comm Servers
Anees Shaikh and Rob Shakir: Openconfig
OpenConfig is an effort amongst many cooperative network operators to define vender-neutral data models for configuring and managing networks programatically.
Craig Partridge: SNMP
The Simple Network Management Protocol, or SNMP, was originally specified in RFC1067, and most recently in RFC1157. The original intent was to make "all IP and TCP implementations be network manageable"—an early form of providing a machine-readable interface so operators could "automate all the things." Craig Partridge played a key role in the early development and standardization of SNMP; he joins us on the History of Networking to discuss the origins and challenges involved in developing SNMP.
Phil Shafer: YANG
YANG is a data modeling language used to model configuration data, state data, Remote Procedure Calls, and notifications for network management protocols,<a href="https://datatracker.ietf.org/doc/rfc7950/"> described in RFC7950.</a> The origins of YANG are rooted in work Phil Shafer did in building an interface system for JUNOS. Phil joins us on this episode of the History of Networking to discuss the history of YANG.
Laura Breeden: FARNT
FARNT was a regional consortium of smaller network operators that eventually helped drive the adoption of TCP/IP and the global Internet, as well as helping efforts to commercialize Internet access. Join Donald Sharp and Russ White as Laura Breeden discusses the origins of FARNT, it's importance in the adoption of early Internet technologies, and the many hurdles regional network operators had to overcome.
Richard Hay: Google Networking
Google fascinates network engineers because of the sheer scale of their operations, and their obvious influence over the way networks are built and operated. In this episode of the History of Networking, Richard Hay joins Donald Sharp and Russ White to talk about some past designs and stories of failure and success in one of the world's largest operating networks.
Keith Mitchell: LINX
LINX is one of the first European Internet Exchanges created. Keith Mitchell joins the History of Networking to talk about the origins of LINX, and the important decisions that shaped its success and the IX community throughout Europe.
Dan Lynch: Interop
Interop is the longest running "show" in the networking space--but it didn't not start as a "show" at all. Dan Lynch, the founder of Interop, joins us at the History of Networking to talk about how Interop really started (hint, it's in the name). One important lesson to learn through this discussion: it is not enough to have standards or open source; in the realm of network protocols, being able to prove multiple vendors can work together is important, too.
Steve Crocker: The RFC Series
Many of the standards that make the Internet "go" are described in Requests for Comments, or RFCs. Steve Crocker joins the history of networking to describe how and why these documents came to be.
David Clark: The IAB and the Early Internet
The Internet Architecture Board "provides long-range technical direction for Internet development, ensuring the Internet continues to grow and evolve as a platform for global communication and innovation." David Clark joins Donald Sharp and Russ White to discuss the origins of the IAB, and relate his experience in the early days of the Internet.
The April 1 RFCs
This episode of the History of Networking is a little different. Because it is the first of April, we have a roundtable of several April 1 RFC authors discussing their work, and a short discussion on the history of the April 1 RFC series. The authors we have on the episode are Donald Eastlake, RFC3092, the Etymology of Foo; Richard Hay, RFC5841, TCP Option to Denote Packet Mood; Carlos Pignataro and Joe Clarke, RFC6593, Service Undiscovery Using Hide-and-Go-Seek for the Domain Pseudonym System; Carlos Pignataro, RFC6592, The Null Packet; and Ross Callon, RFC1925, The Twelve Networking Truths.
David Conrad: APNIC
Each of the seven regional Network Internet Centers (NICs) has a unique origin story reflecting the time in which they were founded, and the operators and regions they represent. David Conrad joins the History of Networking podcast to discuss in the origins of the Asian-Pacific NIC (APNIC) and APRICOT.
Daniel Karrenberg and Mirjam Kuehne: RIPE
In this episode of the History of Networking, Daniel Karrenberg and Mirjam Kuehne join us to discuss their part in the origin of the RIPE NCC, the Regional Internet Registry for Europe, the Middle East and parts of Central Asia. As such, they allocate and register blocks of Internet number resources to Internet service providers (ISPs) and other organisations. RIPE is a not-for-profit organisation that works to support the RIPE (Réseaux IP Européens) community and the wider Internet community. The RIPE NCC membership consists mainly of Internet service providers, telecommunication organisations and large corporations.
Joe Pinto and Phil Remaker: Cisco TAC
The Cisco Technical Assistance Center, or TAC, was as responsible for the growth of computer networking as any technology or other organization. TAC trained the first generation of network engineers, both inside Cisco and out, creating a critical mass of talent that spread out into the networking world, created a new concept of certifications, and set a standard that every other technical support organization has sought to live up to since. Join Joe Pinto, Phil Remaker, Alistair Woodman, Donald Sharp, and Russ White as we dive into the origins of TAC.
Fred Baker: Internet Surveillance
Attempts to create a suite of protocols that made nation-state surveillance existed right from the very beginning of the Internet.
Joel Halpern: Policy Based Management
This recording explores many of the concepts that lie behind intent based networking.
Raj Jain: Congestion Control, Frame Relay, and Early TCP/IP
In this episode of the History of Networking, Raj Jain joins us to talk about his early work with TCP/IP, DECnet, Frame Relay, and congestion control mechanisms. He is the co-inventor of the DEC-bit scheme for congestion avoidance in computer networks which has been adapted for implementation in Frame Relay networks as forward explicit congestion notification (FECN), ATM Networks as Explicit Forward Congestion Indication (EFCI), and TCP/IP networks as Explicit Congestion Notification (ECN). You can find Dr. Jain's current and past work on his page at Washington University.
Peter Jones: NBASE-T
Peter Jones joins the History of Networking to discuss the origins, drivers, and challenges surrounding NBASE-T. While this is a new technology, the work involved in developing the technology and products to make it real reach several years back into the past.
Dan Grossman: ATM
For these two episodes of the History of Networking, Dan Grossman joins Donald and I to discuss the history of Asynchronous Transfer Mode (ATM). While this is a technology that is no longer widely used, it had a major influence on the networking world.
John Chapman: Cable Modems
Before the large cable providers came on the scene, most people accessed the Internet through dial-up MODEMS, connecting to services like America Online, across plain old telephone lines. The entrance of cable providers, and cable MODEMs, allowed the edge of the Internet to explode, causing massive growth. Join Donald Sharp and I on this episode of the History of Networking as John Chapman discusses the origins of the cable MODEM, and the origins of the DOCSIS standards.
Ruzbah Yassani: Cable networks
Fred Baker: Quality of Service and the DS Bit
Quality of service was not a part of the original design of IP, so figuring out how to include QoS in the IP format, and what including QoS would mean, were major steps in the creation of the Internet. Fred Baker played a crucial role in the early QoS work.
Jana Iyengar: QUIC
QUIC is a new transport protocol designed with many of the same features as TCP, but running over UDP.
Daniele Bovio: EARN, RARE, and European Networks
European networks from the mid-1980’s to the late 2000’s underwent a lot of change, bolstered by the rise and fall of America Online, the laying of a lot of subsea cables, and the creation of several organizations, including EARN and RARE, to bolster the spread and use of the Internet. Daniele Bovio joins Donald Sharp and Russ White on this episode of the History of Networking to give us a good overall perspective of this history.
Ivan Pepelnjak: The Internet Behind the Iron Curtain
Ivan Pepelnjak was a founding member of the first IX in Slovenia twenty-five years ago. He joins us to describe the origins of the Internet, from the first dial-up circuits to the founding of the first IX and local DNS services here on the History of Networking. Ivan is an independent consultant and trainer; his work can be found at https://ipspace.net.
Kilnom Chon: An Asian Perspective
Dr. Chon is a professor in South Korea who first encountered IP networking while working in the United States. He carried these technologies back to South Korea, starting the first experiments with this new form of networking, developing native versions of the IP and routing stacks on PDP11's.
Christian O’Flaherty: A Latin American Perspective
The growth of the Internet in Latin America is different than in North America, not least because of the many geographic and political realities in the region. Christian O’Flaherty traces these factors, and how they shaped the Internet in Latin America.
Dawit Bekele: An African Perspective
Dawit Bekele began his journey with the Internet while at college—but on returning to Africa, he discovered there was very little connectivity. While he was not involved in the initial stages of engineering the Internet in Africa, he began as an early user and proponent of connecting his home continent, and is now part of the Internet Society, helping to grow connectivity.
Radia Perlman: Spanning Tree
The spanning tree protocol used to connect Ethernet segments together was actually created after routing, over the span of a single week. Radia Perlman, who invented spanning tree, joins the history of networking.
Dino Farinacci: LISP
Location and identity are often bundled into a single "thing" in IP networking. One attempt to separate these two is the <em>Locater Identifier Separation Protocol,</em> or LISP. Dino Farinacci is the inventor of LISP.
JJ Garcia: The Diffusing Update Algorithm (DUAL)
JJ Garcia is Distinguished Professor of Computer Science and Engineering and Jack Baskin Endowed Chair of Computer Engineering at USC Santa Cruz. He first became involved in packet networks in the 1980's, eventually inventing the Diffusing Update Algorithm (DUAL), which later became the basis for the Enhanced Interior Gateway Protocol (EIGRP). You can find more information about Dr. Garcia at his personal page.
Donnie Savage: EIGRP
EIGRP was, at one time, a dominant routing protocol in large-scale networks.
Saleem Bhatti: ILNP
Mobility problems in networks did not arise with the data center fabric and virtualized load. ILNP was designed to support mobility in IP networks on a campus or global scale.
Juliusz Chroboczek: BABEL
BABEL is a lightweight protocol designed to support ad-hoc networks.
Tony Przygienda: BIER
Supporting service meshes and multicast has always been a challenge on IP networks; BIER provides an answer to these problems.
Alia Atlas: Fast Reroute
The IETF community designed and proposed a few different fast reroute solutions to improve packet switched network convergence.
Adrian Farrel: Path Computation Element
Path Computation Element (PCE) is designed to allow the computation of paths for MPLS and GMPLS Point to Point and
Point to Multi-point Traffic Engineered LSPs. Adrian Farrel, who was involved in the early work on designing an specifying PCE, joins us in this episode of the History of Networking to describe the purposes, process, and challenges involved. You can read more about Adrian on his personal home page, and about PCE on the IETF WG page.
Nick McKeown: P4
Nick McKeown developed the open source P4 language for describing how a switching engine processes packets, allowing engineers to specify and implement customized forwarding to support policy and new protocols in the field.
Bill Yeager: Early PDP11 and Commercialization
Bill Yeager developed the routing code for the PDP11 which ultimately formed the basis for Cisco IOS.
J.R. Rivers: Linux Routing
The open source Linux operating system plays a large role in building routing and host devices; JR Rivers joins the history of networking to discuss the history of routing in Linux.
Sue Hares: GateD Open Source Routing Stack
Sue Hares, cochair of the IDR and I2RS working groups in the IETF, joins Donald Sharp and Russ White to talk about the origins of one of the first open source routing stacks, GateD. Sue was involved in MERIT and the university programs that originated this open source software, and managed its transition to a commercial offering.
Rodney Grimes: BSD and the TCP/IP Stack
BSD is one of the first UNIX implementations, and the IP stack in BSD is one of the first widely used open-source implementations of TCP/IP. Rodney Grimes joins us at the History of Networking to talk about the origins of BSD and these first IP implementations.
Dirk Hohndel: LINUX and SUSE
Started as a consulting company, SUSE was one of the first organizations to begin working in the development and commercialization of LINUX. Through the years, LINUX has become the base for much of the IT world, including many of the open source network operating systems. Dirk Hohndel joins the History of Networking to discuss the origins of SUSE LINUX.
Pamela Dingle: Identity
While identity is not directly a networking technology, it is closely adjacent to networking, and a critical part of the Internet's architecture. In this episode of the History of Networking, Pamela Dingle joins Donald Sharpe and Russ White to discuss the humble beginnings of modern identity systems, including NDS and Streettalk.
Heather Flanagan: Identity
While identity is adjacent to networking, it is an important part of the network engineering world—and is not well understood. Heather Flanagan joins Donald Sharp and Russ White to talk about SAML, unified identity, and some of the practical aspects of verifying a person's identity.
Roland Dobbins: Distributed Denial of Service Attacks
DDoS attacks have not always been as common as they are today, nor have the defensive systems used to protect networks against them. In this recording, Roland Dobbins, who helped to develop and deploy many of the anti-DDoS systems in use today, discusses the history of DDoS attacks.
Martin Casado: Software Defined Networks
Software Defined Networks, or SDNs, were originally designed to solve a specific set of research and deployment problems. While SDN hasn't ever really "taken off" the way many predicted, the concepts and ideas have played a major role in the development of hyperscale and data center fabric control planes.
Jennifer Rexford: Programmable Control Planes
On this episode of the History of Networking, organized through the Association of Computing Machinery, Jennifer Rexford joins Donald Sharp and Russ White to discuss the history of programmable control planes. Dr. Rexford is the Gordon Y. S. Wu Professor in Engineering at Princeton University in New Jersey.
Bob Hinden: IPv6
Bob Hinden is the co-inventor of IPv6.
Doug Comer: TCP/IP
The Transmission Control Protocol, or TCP, is one of the foundational technologies of packet switched networks. TCP not only provides windowed flow control, it also manages the retransmission of data when errors are detected, and sockets for addressing individual applications on a host. Doug Comer was involved in the early development of TCP/IP.
Olivier Bonaventure: M-TCP
MultiPath TCP (MPTCP) is an effort towards enabling the simultaneous use of several IP-addresses/interfaces by a modification of TCP that presents a regular TCP interface to applications, while in fact spreading data across several subflows. Benefits of this include better resource utilization, better throughput and smoother reaction to failures.
You can learn more about M-TCP here and here.
Dave Piscatello: OSI and IP
Dave Farber: The Grandfather of the Internet
Dave Farber advised many of the people who invented technologies fundamental to the creation of the Internet; hence he is called The Grandfather of the Internet.
Vint Cerf: The Father of the Internet
Vint Cerf is considered by many to be the father of the modern Internet.
Larry Landweber: CSNET
Larry Landweber is John P. Morgridge Professor Emeritus of Computer Science at the University of Wisconsin–Madison. He founded the CSNET project in 2979, one of the earliest networks eventually contributing to the creation of the Internet as it exists today. The CSNET eventually became National Science Foundation Network (NSFNET). Larry is credited with making the decision to use TCP/IP on CSNET.
Jon Crowcroft: The Early Internet and Measurement
The early Internet was not only about designing transport protocols, developing control planes, and understanding how to build faster physical transports. Measurement played a huge role in understanding what needed to be changed, what needed to be developed and understanding why the protocols that make the Internet (and other networks) "go" really work. Jon Crowcroft, one of the pioneers in measuring network things, joins this episode of the History of Networking to discuss this history.
Scott Bradner: The Early Internet at Harvard
Scott Bradner was given his first email address in the 1970's, and his workstation was the gateway for all Internet connectivity at Harvard for some time. Join Donald Sharp and Russ White as Scott recounts the early days of networking at Harvard, including the installation of the first Cisco router, the origins of comparative performance testing and Interop, and the origins of the SHOULD, MUST, and MAY as they are used in IETF standards today.
George Sadowsky
George Sadowsky was a pioneer in recognizing the importance of networking technology for economic development, particularly in developing economies. He has worked in over 50 countries to bring training and networking infrastructure to the local population. In this episode of the History of Networking, George recounts some of the early, pre-Internet, work in computer networking, and the development of many of the organizations that make the Internet work today. His web site can be found here.
Dale Finkelson: INET2
The Internet was originally designed as a research network, but eventually morphed into a primarily commercial system. While "Internet 2" sounds like it might be a replacement for the Internet, it was really started as a way to interconnect high speed computing systems for researchers—a goal the Internet doesn't really provide any longer. Dale Finkelson joins Donald Sharp and Russ White for this episode of the History of Networking to discuss the origins of Internet 2.
Dennis Jennings: NSFNET
The NSFNET followed the CSNET, connecting the campuses of several colleges and supercomputing systems with a 56K core in 1986. The NSFNET was the first large-scale implementation of Internet technologies in a complex environment of many independently operated networks, and forced the Internet community to iron out technical issues arising from the rapidly increasing number of computers and address many practical details of operations, management and conformance. The NSF eventually became the "seed" of the commercialized core of the Internet, playing an outsized role in the current design of routing, transport, and other Internet technologies.
Kurt Lindqvist: NETNOD and LINX
Steve Bellovin:USENET
<h2>Mark Nottingham: HTTP</h2>
<a href="https://historyofnetworking.s3.amazonaws.com/http-nottingham.mp3">audio</a>
The original History of Networking Series was produced and distributed through the Network Collective. I am grateful for their support for these recordings.