PING

In this episode of PING, Christian Huitema discusses how looking into the IETF data tracker allowed him to assess "how well we are doing" at document production.

As the IETF has grown, and as the process of developing standards has got more complex its understandable it takes a bit longer to produce a viable RFC but some questions have been made about exactly where in process the delays come from. Are we really doing better or worse than we used to? and, why might that be?

Christian took an interesting approach to the problem, using a random sample of 20 documents from 2018 (initially) and a hand method of collating the issues, and then applied the same methodology back into 2008 and 1998. His approach to measurement was rigorous and careful, separating his own opinions from the underlying data to aide reproducibility.

Christian has a long history of network development and research, with experience in industry, and in the french national computing research institute "INRIA" before joining Bell Communications Research, and Microsoft. He worked on OSI systems, X.500 directories, Satellite communications, and latterly the IPv6 stack including the "Tededo" transition technology, the H/D ratio used in determining IPv6 allocations and assignments in the RIR model, and the QUIC transport layer protocol.

The views expressed by the featured speakers are their own and do not necessarily reflect the views of APNIC.

In this episode of PING, APNIC’s Chief Scientist Geoff Huston discusses the major themes from his recent blog on “Failed Expectations”

In a trip down memory lane, the podcast ranges over the 40 year plus history of how we came to have the current Internet as we know it, and some of the “road not taken” alternates which were under consideration at the time. In this context. “Failed” doesn’t have to mean “failed to work” -it can mean the technology simply wasn’t chosen, or it can be the “failure” to turn off something which was believed to be at best temporary!

In part, the story of IPv6 deployment is part of this mismatch of expectations and reality, because nobody sought the outcome we’re now living through, of a 20 plus year transition from 32 bit addresses to a world of 128 bit addressing. IPv6 was designed with an eye to the needs of addressing at scale, but the emergence of a transfer model, and continued improvement in NAT (and deployment of Carrier-grade NAT or CGN) at scale, worldwide has perpetuated a 32 bit address and routing world. IPv4 Internet is the “little network which could” and refuses to go away quietly.

The views expressed by the featured speakers are their own and do not necessarily reflect the views of APNIC.

In this episode of PING, Verisign Fellow Duane Wessels discusses notable changes in the DNS root zone over the last 13 years.

Duane joined Verisign in the early stages of DNSSEC deployment and has conducted measurements of DNS for many years, in his measurement factory days, and in DNS OARC as well as inside Verisign.

The significant changes to the DNS root zone, and it's implications for the root zone operators are discussed: Deploying DNSSEC, the first DNSSEC KSK key changes, the increase in packet sizes with RSA keylength changes, and the future KSK and ZSK algorithm changes

Read more about DNS and DNSSEC on the APNIC Blog.

Here's some articles from the blog which discuss the issues:

The views expressed by the featured speakers are their own and do not necessarily reflect the views of APNIC.

In this episode of PING, APNIC's Chief Scientist Geoff Huston discusses the question of buffers, flow control and 'efficient' use of a network link.

How do we maximise the use of a given network path, without knowing everything about its size along the way? It turns out, the story isn't as simple as "more is better" because sometimes, adding more memory to the system adds delay. Modern TCP's flow control algorithms are being modified to react to delay as well as loss, and become more efficient at occupying the available space. At the same time bit-marks inside the IP packet are modifying how end hosts can react to signals of congestion along the path. Are these two mechanisms in conflict? how do they stack up, and achieve critical mass in deployment?

Read more about TCP and flow control on the APNIC Blog.

Here's some articles from the blog which discuss the issues:

• Comparing TCP and QUIC (Geoff Huston)
• Does TCP keep pace with QUIC? (Konrad Wolsing)
• TCP Congestion Signatures (Srikanth Sundaresan)
• Striking a balance between bufferbloat and TCP queue oscillation (Ulrich Speidel)
• TCP initial window configurations in the wild (Jan Rüth)
• Underload: The future of congestion control (Safiqul Islam)
• Beyond bufferbloat: End-to-end congestion control cannot avoid latency spikes (Bjørn Teigen)
• Congestion Control at IETF 110 (Geoff Huston)

The views expressed by the featured speakers are their own and do not necessarily reflect the views of APNIC.

In this episode of PING, Dr Romain Fontugne, the deputy director of research at IIJ Labs in Tokyo discusses the IIJ "Internet Health Report" and AS Hegemony (or network centrality) in particular.

This is a data model they have been working on for some time (6 years now) which exposes dependencies between ASs in BGP, both directly (as in customer-cone) and indirectly through transitive dependencies. It's a fascinating insight into how BGP dependencies can be seen through the state of the routing table worldwide, and how IIJ are helping BGP speakers understand the dependencies in the transit paths they use. It's also a fertile space for student engagement with google summer of code opportunities.

You can read more about AS Hegemony on the APNIC blog, as well as other posts Romain has made reflecting IIJ's research.

• AS hegemony: measuring AS interdependence (2018)
• BGP zombies (2019)
• The impact of COVID-19 on last-mile latency (2020)

IIJ's Internet Health Report (IHR) is reachable at http://ihr.iijlab.net 

In this episode of PING, APNIC's Chief Scientist Geoff Huston discusses Low Earth Orbit (LEO) satellite communications and the amazing effects they are having on internet reach worldwide.

Read more about the issues in LEO and satellite communications on the APNIC Blog.

Here's some recent articles of note:

• Getting hands-on experience with Starlink (Ulrich Speidel)
• Everything, everywhere, all the time (for the internet at least) (George Michaelson)
• The APRICOT Panel discussing satellite broadband in the Asia Pacific region (Geoff Huston, Dan York, Debopam Batterchee, Ulrich Speidel, Mike Puchol)
• How does starlink compare to broadband (Mohamed Kassem in 2022)
• Fact checking Starlink’s performance figures (Martino Trevisan in 2022)

The views expressed by the featured speakers are their own and do not necessarily reflect the views of APNIC.

In this episode of PING, Dr Rolf Winter, the Professor of Data Communications at Augsburg University of Applied Sciences discusses his work on ‘reverse traceroute’, which is an approach to using the well-known traceroute mechanism but driven from the other end.

The inherent problem with traceroute and its related diagnostics is that it only informs you about the path outwards from your address to the other end.

Reverse traceroute is an attempt to ‘mechanize’ the reverse path information, using proposed new codepoints in the Internet Control Message Protocol (ICMP). Rolf discusses this approach and some of the logistical issues with attempting to modify an established protocol like ICMP, and measurements of the acceptability of proposed new codepoints in the wild.

Read more about Professor Winter’s work on the APNIC Blog:

• Troubleshooting ‘the other half’
• Watch his presentation at DENOG 14
• Visit his GitHub code repository

In this episode of PING, APNIC's Chief Scientist Geoff Huston discusses DNSSEC and presents a case "for" and "against" deployment, in the context of complexity, fragility, and impact on the DNS process at large. DNSSEC is net beneficial but its by no means automatic to deploy it protecting a zone.

Read more about the issues in DNSSEC deployment on the APNIC Blog:

• to DNSSEC or not? (Geoff Huston)
• IP Fragmentation and the DNS - Mitigation (Carsten Strotmann)
• DNSSEC Validation - Performance Killer (Petr Špaček)
• [Podcast] DNS spoofing is a non-issue if we all do DNSSEC (Geoff Huston)
• Addressing the challenges of modern DNS (Moritz Müller)

And for Geoff's continuing measurement of DNSSEC see https://stats.labs.apnic.net/dnssec for his daily updated measurements of DNSSEC validation.

The views expressed by the featured speakers are their own and do not necessarily reflect the views of APNIC.

In this episode of PING, Andre Geldeblom from APNIC product development discusses how APNIC is measuring user experience, satisfaction and engagement with the "Orbit" system we deployed to provide web services integrated with email.

Andre discusses the different motivations and mechanisms we're using at APNIC to understand "UX" and how this integrates in our planning to deliver "value for money" to the APNIC community.

Read about the different motivations and mechanisms we’re using at APNIC to understand “UX” and how this integrates in our planning to deliver “value for money” to the APNIC community.

• Upcoming changes to MyAPNIC (Andre, October 2022)
• how APNIC is implementing your feedback (Dale, September 2022)
• Improving Accessibility at APNIC (Dale, August 2021)
• APNIC help centre now live (Andre, February 2021)
• Help APNIC create better products and services (Dale, September 2020)

The views expressed by the featured speakers are their own and do not necessarily reflect the views of APNIC.

In this episode of PING, APNIC's Chief Scientist Geoff Huston discusses the current situation in BGP across IPv4 and IPv6. Historically. we've met the "running out of memory" problem with incremental upgrades but things have now come to a place where "simply adding more memory" may not be the answer, and the dynamics of BGP growth appear to have changed.

Geoff explores the changing surface of the BGP default-free zone, and what it means for routing technology and the ISP

Read more about the APNIC Labs analysis of BGP on the APNIC Blog:

• BGP in 2022 - The Routing Table
• BGP in 2022 - BGP Updates
• IP Addressing through 2022

And for Geoff's continuing measurement of BGP see bgp.potaroo.net for his daily updated history of BGP from 1994.

The views expressed by the featured speakers are their own and do not necessarily reflect the views of APNIC.