NetUK3

6 Jul 2026, 09:00 → 7 Jul 2026, 17:00 Europe/London

The Porter Tun (1st Floor) (The Brewery)

NetUK3 took place on 6-7 July 2026 in London

---

NetUK is an open and vibrant community of individuals and organisations that are passionate about network infrastructure and wider internet technologies across the United Kingdom. Its events enable participants to share key technical knowledge, discover community achievements, develop industry connections and promote participation regardless of experience or background.

NetUK 3 - Sponsorship 2026.pdf

Monday 6 July

Workshop Registration

Workshops

1 Workshop: An Introduction to Data Centre Networks

Given the rapid pace at which technology is evolving, there is a growing gap between fresh graduates and the expectations of entry- to mid-level roles. Many positions that previously required highly specialised expertise now demand a broader and more diverse skill set. At the same time, data centres networks have rapidly evolved for almost every type of organisation—including enterprises, colocation and service providers. Changes in business requirements, traffic types, patterns, applications and increased availability has redefined data centre architectures. 

In this context, we would like to propose a workshop that introduces the key concepts behind modern data centre networks and campus fabrics. The session would explore the inner workings of modern data centre fabrics, how interconnections extend across different environments, and the role telecommunications providers play within this ecosystem. We would also provide a brief overview of common tools and traffic flows. As data centres are now classified as Critical National Infrastructure (CNI) in the UK, we believe it is important to highlight their real-world importance and the need for designing highly available, high uptime and highly resilient architectures. As part of the workshop, we would present three practical real-world use cases to demonstrate how end-to-end networks operate in modern data centre environments. These examples would help illustrate the critical nature of these mission-critical systems and provide attendees with a clearer understanding of the broader picture.

We would also briefly discuss the evolution of modern data centre networking, giving participants insight into how architectures have developed to support today’s scale, resilience, and performance requirements. To make the workshop engaging and interactive, we propose having two speakers, each covering a different dimension of the topic. Three real world case studies would further help tie theoretical concepts with real life applications for these architectures. These examples would help illustrate the critical nature of these mission-critical systems and provide attendees with a clearer understanding of the broader picture. We would also actively involve the audience throughout the session to ensure participants remain engaged and enthusiastic.  There are no prerequisite requirements for the audience and we believe our talk will cover audiences with limited experience to computer networking as well as entry-level to mid-level network engineers.

Speakers: Mr Foeh Mannay (Xantaro), Utkarsh Shah (Xantaro)

NetUK - An Introudction to Data Centre Networks - Workshop - Xantaro - Foeh Mannay & Utkarsh Shah - Final - 8th July 2026.pptx

2 Workshop: Open Source Data Management Platform - Infrahub

Open-Source Infrastructure Data Management : Infrahub Workshop

Infrahub reimagines infrastructure automation by combining Git-like version control capabilities on top of a graph database. The platform's three pillars - flexible schema, native version control, and unified storage - create an ideal teaching environment where participants can experiment with real-world scenarios in this hands-on workshop.

Module 1: Foundation (40 minutes) establishes core concepts. Participants begin with Infrahub's sandbox environment, creating their first network objects and understanding the relationship between schema, data, and version control. This foundation addresses the immediate need for practical source of truth implementation.

Lab 1: Network Source of Truth Setup (20 minutes) introduces participants to Infrahub's interface through basic device and IP address management tasks. Beginners focus on GUI navigation while advanced users explore CSV imports and API interactions. This foundational lab establishes confidence and familiarity with core concepts.

Module 2: Config Generation (40 minutes) builds on foundational knowledge by connecting Infrahub to existing automation tools. After covering concepts around GraphQL queries and Jinja2 templating, we will cover generating device configurations.

Lab 2: GraphQL Mastery (20 minutes) transforms participants from consumers to creators of network data queries. Using the built-in GraphiQL explorer, attendees progress from simple device queries to complex relationship traversals, preparing them for real-world data extraction and integration scenarios.

Speakers: Alex Gittings (OpsMill), Pete Crocker (OpsMill)

Doors Open, Registration & Lunch: In-person attendees

Newcomer Introduction

NetUK3 Plenary

3 NetUK Welcome

Welcome to NetUK3!

Speaker: Dave Pumford (PTX Technologies / NetUK)

NetUK3 - Welcome to NetUK3 - Day1 - v2.pptx

4 When UK Traffic Takes an Unplanned Detour: Routing Security for the UK Peering Community

In June 2019, a Swiss data center accidentally leaked over 70,000 routes to China Telecom in Frankfurt. Within minutes, China Telecom announced those routes globally, and for over two hours, massive volumes of European mobile traffic — destined for networks in the Netherlands, Switzerland, and France — were silently rerouted through Chinese state-controlled infrastructure. In the summer of 2025, a BGP hijack disrupted DNS root server infrastructure, shaking the foundation of global name resolution. And in between, researchers documented a new class of "stealthy" BGP hijacks where a network's RPKI checks all pass and the BGP table looks perfectly clean, yet traffic is being diverted through non-validating neighbors without the victim ever knowing.

These are not abstract, far-away problems. The United Kingdom sits at the heart of global Internet interconnection. LINX is one of the largest Internet Exchange Points in the world. Traffic traversing UK peering fabrics touches networks across Europe, the Middle East, Africa, and beyond. When a route leak or hijack occurs anywhere on the planet, UK operators and their peers are in the blast radius — and when it happens to a European neighbor, the impact is immediate.

The good news: the UK peering community has made real progress. LINX and LONAP both enforce RPKI validation on their route servers, filtering invalid routes before they reach members. But route server filtering is only the first line of defense. Members peering bilaterally — without the route server as an intermediary — remain exposed unless they deploy validation themselves. And even networks with full ROV enabled face blind spots: route leaks that ROV cannot detect, path manipulation attacks that require ASPA to prevent, and stealthy hijacks that demand observability tools like BMP to even notice.

This presentation provides a practical, UK-focused roadmap for closing these gaps. We examine what LINX and LONAP route server filtering already protects against and where the remaining exposure lies for bilateral peers. We introduce ASPA — the emerging RPKI-based path validation standard that all five Regional Internet Registries have committed to supporting by the end of 2026 — and explain what it means for UK operators signing their first ASPA objects. And we make the case for BMP as the missing security observability layer: the tool that turns filtered invalid routes from silent discards into active threat intelligence, enabling operators to detect anomalies that neither ROV nor ASPA can surface on their own. Attendees will leave with a clear, prioritized action plan — from quick wins they can implement this quarter to strategic investments that will keep UK peering infrastructure resilient against the next generation of routing attacks.

Speaker: Ritesh Mukherjee (Nokia)

Ritesh NetUK 2026.pptx

5 Understanding and optimising transceiver efficiency using internal metrics for improved power savings

In high-performance optical communication systems transceiver health and efficiency are critical to network reliability and energy consumption. This presentation explores the powerful capabilities of Versatile Diagnostics Monitoring (VDM) features found in modern optical transceivers (beyond the speed of 100G) with a particular focus on the Thermoelectric Cooler (TEC) current metrics. By analysing TEC current alongside temperature and cable length data we should be able to identify the optimal operating conditions that will minimise power consumption while maintaining performance. In order to prove this thesis we developed a couple of software prototypes to perform the data analysis. Using real-world VDM data this talk will demonstrate how to evaluate and visualise transceiver efficiency in terms of Watts, uncovering practical insights for engineers aiming to design or operate greener, more efficient optical networks and finally save energy.

Speaker: Dr Gerhard Stein (Flexoptix)

Flexoptix - TransceiverPowerConsumption_NetUK.pdf

6 Data Signal-Based Distributed Fibre Optic Sensing

Physical fibre cable problems cause roughly a third of all network outages. Reducing Mean Time to Repair (MTTR) isn’t just a technical goal-it’s an economic necessity. While Distributed Fibre Optic Sensing (DFOS) can solve this, the high cost of deploying dedicated sensing fibres or extra lasers remains a major obstacle for widespread operator adoption.

To solve this, our joint R&D initiative between Bangor University’s DSP Centre of Excellence, Sorrento, and Comtek Networks has focused on developing a breakthrough "Data-Driven Sensing" architecture. By enabling the live, active data-carrying signal itself to act as the sensor, we classify and localise fibre faults with ~98% accuracy, all with no extra fibre assets, no additional lasers, and no downtime.

This matters because it eliminates the massive capital overhead of traditional standalone DAS monitoring. It allows operators to protect critical backbone routes and detect "near-miss" physical threats using the active, revenue-generating fibre you already have in the ground. Finally, the presentation outlines a forward-looking research roadmap for developing this technology to simultaneously monitor the physical infrastructure running alongside most fibre networks.

Speaker: Jasmine Parkes (Comtek Network Systems and Bangor University)

netuk2026v2.pdf

netuk2026v2.pptx

Sponsor Content

7 Agentic AI Network Automation

Megaport will be presenting the move from manually depoying NaaS services via the Megaport portal to using MCP and Agentic AI to manage customer infrastructure.

Speaker: Gary Taylor (Megaport)

Megaport - NETUK3 July 26.pdf

Megaport - NETUK - July 26.pptx

NetUK3 Plenary

8 Hollow Core Fibre - Nothing is better than glass for guiding light

The impressive advantages when using light not silicon make the promise of HCF something worth taking notice of.
This presentation takes a short look at the technology behind HCF, outlines the benefits and possible applications whilst highlighting some constraints (& solutions) in market dominated with singlemode fibre & LC/APC connections

Speaker: Steve Jones (Huber+Suhner)

260626 H+S Hollow-Core Fiber.pptx

15:20 30 Minute Break

NetUK3 Plenary

9 Data Centers in Space - The Big Picture

Data Centers in Space
Geoff Bennett
Director, Solutions and Technology, Nokia

The race for companies to develop the next best Artificial Intelligence model means that they need to deploy processing capacity at unprecedented scale. We are already seeing the construction of multi-gigawatt scale data centers, and this creates significant challenges to find sufficient space, power and fresh water for these facilities.
In addition, the sheer scale of such data centers brings them into direct conflict with local human populations, leading to an increasing backlash against the permitting of these projects in populated areas.
Beginning with remarks by Bill Gates as far back as 2014, several technology leaders have outlined plans for massive constellations of orbiting data centers. In January this year SpaceX filed plans with the US Federal Communications Commission for up to one million “orbital data scenter” satellites. Google, with its Project Suncatcher, is suggesting that space-based data centers could be a reality as soon as 2027. And multiple startups are already raising funds to develop the building blocks of these new technologies.
But how realistic is the goal itself? What are the major hurdles in our way, and how easy is it to get over or around them? Are there some problems that are just insoluble? And how much damage are we likely to inflict on humanity as we chase this dream?
I begin with a simple question - what problems is this idea of space-based data centers trying to solve? From there I will look at the technologies we need to develop, the scale of what would have to happen, and the unavoidable consequences of filling up space in this way.

Speaker: Geoff Bennett (Nokia)

2026-04 Data Centers in Space - NetUK3 .pptx

10 Telecoms Inventory for GPON using Kuwaiba

Kuwaiba is an open source telecoms inventory solution which originated in South America but is increasingly gathering attention from service providers globally. It provides the power of many commercial systems which model network physical and logical resources and services at a significantly lower cost.
Recently a group of students from Solent University did an industrial consulting project to evaluate how easy it would be to design and model a city wide GPON network in Kuwaiba using data imported from UK UPRNs and OpenStreetmap GIS information.
This talk will provide a fascinating introduction to what the Kuwaiba project is and how it could be adopted as a tool by ISPs and service providers in the UK.

Speakers: Mr Corey Maltby (Solent University), Craig Gallen (Entimoss and Solent University)

kuwaibaImporter1.mp4

Netuk-Inventory-gpon-kuwaiba_V_1.0.pdf

Netuk-Inventory-gpon-kuwaiba_v_1.0.pptx

11 Edge Network Evolution

This short talk outlines how Meta's edge infrastructure is evolving to meet the demands of changing workloads — from traditional content delivery to emerging AI-driven experiences. The speaker explains how Meta's network topology is being reshaped to solve for both efficiency and quality of experience, ensuring that networks of all sizes continue to have access to direct, 1:1 peering relationships with Meta. The speaker demonstrates how these architectural shifts help connect people with Meta's apps and services while landing optimisations that enable operators to defer or avoid capital expenditure and deliver quality improvements for ISPs

Speaker: Ben Ryall (Meta)

Meta @ NetUK3 (6-7th July 2026).pdf

12 Which in my switch? LPO, CPO, XPO optics. What are they, when to use

I will first detail what LPO (linear-drive pluggable optics) are versus traditional pluggables, and contrast CPO (co-packaged optics) with this, and their respective practical benefits such as cost, latency, power. I will talk about switch/routing use-cases where LPO wins over traditional pluggables today and vice versa - plus how this will evolve as port data-rates increase. Then I will talk about possible future evolution to much higher density with CPO and the benefits and challenges this brings, plus the alternative approach of XPO (extra-dense pluggable optics).

Speaker: Dr Robert Friskney (Arista)

2026-06-18 LPO CPO optics for NetUK3 v4.pptx

2026-06-18 LPO CPO optics for NetUK3 v4 - Read-Only.pdf

Social Event

Tuesday 7 July

Doors Open & Registration: In-person attendees

NetUK3 Plenary

13 LON2 Network Refresh

LON2 was originally launched in 2002 following LON1’s increasing popularity and was created to ensure there was no critical single point of failure in the UK’s internet connectivity. LINX’s dual LAN infrastructure design in London has been something that sets us apart over the years from other Internet Exchange Points. Operating a redundant, diverse and resilient solution in house for members means not having to look elsewhere for this critical capability. In 2018 LINX announced that they were the world’s first IXP to deploy a disaggregated network using EVPN, with hardware and software sourced from alternative vendors.

However, with end of life equipment, how did LINX refresh a LAN with 1/3 of its membership on without issues... or was there?

Speaker: Jan Kayser (LINX)

LON2_V12.pptx

14 Everything is Fine and Our Automation Works Perfectly

...as long as you update your IRR, give us RTBH fields, and fill out this survey.

Speaker: Stefan Funke (Inter.link)

NetUK3-Stefan-Funke.pdf

15 SCION: Secure Path-Aware Internet Routing for Critical Infrastructures

The Internet routing system (BGP) is rarely considered when discussing cybersecurity even though this remains an insecure aspect of the network and is increasingly used as an attack vector. Even though measures such as RPKI have been developed in recent years to improve routing security, adoption has been limited and still do not address the fundamental issue of limited control over where data is transiting.

SCION is a secure path-aware Internet architecture designed to address many of the limitations of BGP by offering high resilience to routing attacks and path selection for Internet users and operators. The ability to use Internet connections from different ISPs can be more flexible and cost effective than relying on dedicated point-to-point links, but the commodity Internet is not able to provide sufficiently trustworthy, reliable and resilient connections for many critical infrastructures, especially those with safety critical traffic. Having the ability to control the paths that traffic is sent through intermediate networks, is also important for reducing the possibilities for traffic interception, reconnaissance, man-in-the-middle and denial-of-service attacks, as well as ensuring that data can be geofenced especially looking towards a post-quantum world.

SCION addresses many of the existing issues with how traffic is sent over the Internet, including the issues of network trust, path selection, fast failover, and control over where data is transiting. It implements a trust model based around logical groupings of networks sharing a common jurisdiction with agreed trust policies and a standalone trust root (i.e. one that is not reliant on third-party Certificate Authorities) to validate the networks. This ensures that traffic is only exchanged between networks that are members of a trust domain, except that traffic which is explicitly allowed to be exchanged with other trust domains and the rest of the Internet. In particular, this allows truly sovereign networks to be built and operated.

SCION has commercial and open-source implementations and is already in production use in the Swiss financial services, healthcare and power utility industries, as well as being evaluated by other sectors (e.g. defence) running critical infrastructures. It is also being developed as an vendor-neutral IETF specification.

This proposed talk will discuss the use cases and the SCION design, architecture and trust mode. It will also discuss how SCION is deployed and how it compares with other current and proposed solutions such as RPKI, MPLS, SD-WAN, Segment Routing and BGPSEC.

Speaker: Kevin Meynell (SCION Association)

SCION (NetUK 3) for presentation.pptx

SCION (NetUK 3).pdf

16 Engaging with PeeringDB's Product Committee

PeeringDB is both a directory of interconnection opportunities and an API enabling automation. Decisions on which features should be developed, improved, or changed sits with its volunteer Product Committee. But the Product Committee relies on PeeringDB's users to inform its decisions.

This presentation will explain how and why to engage with the Product Committee. Attendees will then be asked to discuss how or why they'd use a feature or change currently being discussed.

The specific issue to be discussed will be decided nearer the time as it is several months in the future.

Speaker: Leo Vegoda (PeeringDB)

20260706_NetUK3_Leo_Vegoda.pptx

20260706_NetUK3_Leo_Vegoda.pptx.pdf

11:15 30 Minute Break

NetUK3 Plenary

17 Rethinking Network Monitoring

Traditional network monitoring — SNMP polling, static dashboards — tells us what broke, not why. This talk reframes monitoring as an observability problem and walks through a modular, open-source stack enriched with a source of truth. We'll cover correlating metrics, logs and flows, event-driven automation, and AI-assisted root cause analysis — so you leave with a practical blueprint you can pilot yourself.

Speaker: Mr David Flores

Rethinking Network Monitoring (NetUK).pptx

18 Autonomous System Provider Authorisation

ASPA stands for Autonomous System Provider Authorisation and is a relatively new type of object standardised in the RPKI ecosystem. ASPA objects are signed by AS owners and allow an AS to declare its authorised upstream providers, which prevents route leaks, BGP path spoofing, and valley-free routing violations. They can be created in the hosted RPKI systems of ARIN and RIPE NCC, or using delegated RPKI CA software such as Krill. ASPA standard has already seen significant uptake and is supported by some of the software routers.

Speaker: Ralph Smit (RIPE NCC)

NetUK3-ASPA.pdf

19 The Network Automation Forum Framework: A vendor-neutral reference model for building automation at scale

The Network Automation Forum spent the last year developing a common reference architecture that network engineers, vendors, consultants, and end-users can actually use.

This talk walks through the framework architecture, explains why we needed standardised language around network automation, and shows how you can apply it whether you're starting from scratch or trying to make sense of what you already built.

Reference: https://reference.networkautomation.forum/Framework/Framework/#presentation

Speaker: Pete Crocker (OpsMill)

NetUK Talk - NAF Framework 2026-07.pdf

13:15 Lunch Break

Sponsor Content

20 The Great Escape from On-Prem

Modernising infrastructure isn’t simply about moving workloads to the cloud; it can also demand rethinking how foundational network services are delivered, operated, and secured.

In this session, we’ll share a practical architectural pattern for transitioning from traditional on-premises deployments to a Protocol-as-a-Service model. We’ll cover migration strategy, operational transformation, and integration with modern security architectures such as SASE.

Speaker: Ed Daniel (Infoblox)

NetUK3_Infoblox_2026_FINAL_v2.pptx

NetUK3 Plenary

21 Building a National NOG Community

netUK's predecessor organisation, the United Kingdom Network Operators' Forum (UKNOF), ran over 50 conferences and operated community platforms for British Internet infrastructure providers and professionals between 2005 and 2023.

This talk, by UKNOF's founder, is a retrospective of that time, from UKNOF's early volunteer high-energy ad-hoc formation, through its growth and success, to one of the largest national Network Operator Groups (NOGs) globally, by way of pandemic adaptions, and latterly sustainability challenges.

As well as looking at some high-note presentations, tools, fun anecdotes and various numbers, the theme of professional development through community building runs through everything that UKNOF did. Another important thread is openness of participation to all, in the true spirit of international Internet bodies such as RIPE and IETF.

There will also be reflections on what worked, and what didn't, including the benefits and limitations of a strongly pragmatic approach to governance, together with some lessons learned for the long-term sustainability of nonprofit entities in the Internet operations space.

Speaker: Mx Keith Mitchell

Building a National NOG Community

netUK3-UKNOF.odp

22 Operators Checklist for the October 11th DNSSEC KSK Rollover

This presentation explains what will happen on October 11th, and how to make sure your validating resolvers are up to date and ready for the event.

The attached slide deck is based on Roy Arends from ICANN and is used with his permission. I will present his material, with a couple of modifications of my own.

Speaker: David Groves (Nominet)

key-roll-netuk3.pptx

15:15 30 Minute Break

NetUK3 Plenary

23 NetUK Update

Updates about NetUK including the finances

Speaker: Nick Wenban-Smith (Nominet)

NetUK3 - MC Update - Day2 - v2.pptx

24 Evaluation of BGP Roles (RFC 9234) with the aid of an AI Assistant

LLMs have greatly improved recently with the advent of reasoning models. We will set the scene by showing the results of LLM benchmarking tests against networking certification exams. We will then show an example in which an LLM, in conjunction with a Junos MCP server, configures, tests and documents BGP Roles (RFC 9234).

Speaker: Dr Julian Lucek (HPE)

NETUK-July26-JLUCEK-v3.pdf

NETUK-July26-JLUCEK-v3.pptx

slide28.mp4

slide9.mov