BTW.Media
Daily editionStrategic Internet Intelligence
Sign InRegister

Live Wire

View all briefings
Oregon Democrats plan three-year data-centre moratorium|Sandisk’s datacenter surge still carries a NAND price tag|Liquid’s Lagos optical links trade trenching time for line-of-sight discipline|FCC reportedly drafts ban on new Chinese optical transceivers|Paul Abfalter appointed FLAG chief executive|Allentown has turned a data-centre setback into an evidence ledger|Telecentro's AS27747 Identifies a Network, Not Its Resilience|Blue Origin has named New Glenn’s failed valve; proving the fleet fix is the harder step|Private Relay’s blind spot begins where Safari no longer owns the request|When an AI safety test crosses into the real internet, authorisation becomes the safety system|SpaceX’s AI build-out has a conversion problem: turn $15.8bn of quarterly capex into durable demand|Zenith’s contact-data breach moves risk from the vault to the inbox

Lead Report / Asia-Pacific National Telecom

An unbranded network operator reviews a rollback checklist beside displays showing a redundant emergency-call routing path.
An Australian mobile-network change interrupted voice service for 80 minutes, but the more revealing failure came after engineers began the rollback: the organisation knew it had a significant outage at 1:22am and did not notify the national emergency-call operator until 9:07am. The incident shows why continuity depends on evidence, handoffs and tested operating controls, not simply on a network diagram that includes backup paths.

TPG's 2024 voice outage turned emergency-call notification into a change-control test

An Australian mobile-network change interrupted voice service for 80 minutes, but the more revealing failure came after engineers began the rollback: the organisation knew it had a significant outage at 1:22am and did not notify the national emergency-call operator until 9:07am. The incident shows why continuity depends on evidence, handoffs and tested operating controls, not simply on a network diagram that includes backup paths.

Read Full AnalysisAsia-Pacific National Telecom Intelligence
  1. Abstract control-plane continuity diagram showing edge nodes, a facility dependency break, and recovery paths.

    Global Cloud Services

    Cloudflare's 2023 PDX outage made control-plane disaster recovery an accountability test

    On November 2, 2023, Cloudflare's edge network continued to serve traffic while its control plane and analytics services were disrupted. The difference matters. A data plane is the part of a network that keeps packets and requests moving; a control plane is where operators and customers change configuration, see state, and recover from incidents. Cloudflare's postmortem says the chain began with a power event at PDX-04, a data center in Oregon, and exposed hidden dependencies in services that were supposed to survive a facility failure. The accountability test is not whether a large edge can keep forwarding. It is whether the operator can still configure, observe, restart, and prove recovery when the place hosting part of its control system fails.

  2. Two telecom operations engineers review an abstract upload and error-feedback path beside unbranded network equipment.

    Asia-Pacific National Telecom

    Aussie Broadband's 2022 IPND failures showed why an emergency database needs a working feedback loop

    An emergency database can receive files every day and still drift away from reality if nobody checks what came back. In 2022, Australia's communications regulator found that Aussie Broadband had not supplied or updated required public-number records on time, while an error-reporting problem also kept feedback files from reaching the people meant to act on them. The case is not evidence that every affected record caused an emergency failure. It is a practical lesson about closing the loop between a live phone service and the ledger used by public-safety systems.

  3. Two emergency call-centre operators monitor routing screens and review a continuity contact binder in a pre-dawn operations room.

    Asia-Pacific National Telecom

    Telstra's March 2024 Triple Zero disruption showed that a backup must preserve the call, destination and location

    Australia's communications regulator found 473 breaches after a technical disruption at Telstra's Triple Zero call centre. The event was not simply a server outage. It exposed three different continuity duties that are easy to collapse into one: keeping the caller connected while a call is transferred, reaching the correct emergency service organisation, and carrying forward the caller information already available to the emergency-call platform.

Member Access

Daily Strategic Internet Intelligence, Free for Members

Become a BTW member for free daily updates across governance, market, membership, and RIR Watchdog sessions. Join now and manage your delivery preferences anytime.

  • 01Daily intelligence brief
  • 02Governance + market + RIR
  • 03Preference control anytime
Create AccountSign In

BTW Media on Air

Selected briefings, interviews, and live sessions from the BTW intelligence team, focused on internet infrastructure, governance, and market signals.

View All

Twilio Reported Partial Recovery Across Safaricom SMS Routes

Nine Michigan Water Systems Reported Malicious Cyber Activity

Zoom Stayed Operational While Meeting Quality Declined

Twilio Saw Safaricom Delivery Receipts Recover After 57 Minutes

Fiber57’s Routing Silence Leaves Pimentel’s Recovery Path Unproven

Rolls-Royce Data-Centre Growth: Orders, Revenue and Margin Are Not Equal

China’s Data-Centre Expansion Continues Despite Overcapacity Risk

India’s 12 GW Data-Centre Forecast Still Faces the Delivery Test

Exelon’s 11 GW Queue: Which Data-Centre Projects Are Real?

How a DNSSEC Rollover Disrupted Germany’s .de Domains

InterNetX Anycast: Global Visibility Does Not Prove Independence

JLL and LASALLE: Two Brand Domains, One Continuity Test

AS134772: Dongguan’s Dual-Stack Routes Are Visible—But Is Service?

SHANGRILA’s Dual Namespace: Brand Control or Real Resilience?

AMS-IX’s LACP Leak: Multiple Links, One Shared Failure

A network operator reviews two abstract verification paths beside an unbranded phone, SIM tray and security token.

Market / Companies

Optus's Coles Mobile porting breach showed that a phone number is a control record, not just an account field

Australia's communications regulator found that Optus, acting as the gaining provider for the Coles Mobile brand, proceeded with 44 unauthorised mobile-number ports without completing the required additional identity verification. The incident is easiest to understand when a phone number is treated as a unique network resource whose operational control is being transferred—not as a harmless text field being copied from one customer account to another.

A network standards contributor, seen from behind, compares unlabeled network-model diagrams across two monitors.

Market

Rob Shakir and the Work of Describing Networks Across Vendor Boundaries

A large network rarely fails because its engineers lack a name for a router. It fails in harder places: when two devices use different names for the same kind of state, when an automation system reads one vendor correctly and another incorrectly, when a route moves between tables without the intended policy, or when a replacement team cannot reconstruct what an old configuration was meant to do. These are not glamorous problems. They are problems of description, transfer, and restraint.

A photorealistic primary and standby network facility linked by redundant fiber, with the transfer path paused at a clearly separated mechanical authorization gate.

Governance

The NRO's AFRINIC Contingency Plan and the Missing Operator Consent

The regional registries are right to prepare technical and financial support for an AFRINIC failure. But mirrored records, peer expertise and emergency funding do not by themselves authorize another registry to inherit contracts, exercise resource-management discretion or bind the operators that depend on AFRINIC. The unresolved issue is mandate, not the value of backup.

Photorealistic editorial portrait of Carlos Martinez-Cagnazzo in a routing-security operations setting.

Market / Trends

Carlos Martinez-Cagnazzo and the Boundary Work of Routing Security

Carlos Martinez-Cagnazzo's public record does not describe a lone architect making the internet safe. It shows a more instructive kind of authority: a regional registry strategist working where cryptographic standards, trust-anchor operations, operator education and cross-border incident coordination meet, and where each technical control remains dependent on institutions and people using it well.

Anonymous conference entrants move past an aggregate counting grid as four paths toward speaking, volunteering at a help desk, mentoring and committee service break at missing data connections.

Governance / NOGS

Counting Entry, Not Continuity: Where NANOG's Participation Record Stops

NANOG’s annual reports have become increasingly detailed about how people appear in its meeting record: registrations by mode, first-time participants, membership status, students, speakers, women in attendance, employment bands and organisation types. Its programme record also describes orientation, mentoring, ambassadors, a help desk and invitations to take on later roles. What the reviewed public record does not publish is the bridge between those two views—the share of a first-time cohort that returns, speaks, volunteers, becomes a member or serves in a formal role. That is a bounded measurement gap, not evidence that newcomers disappear.

Editorial position

Why BTW Media exists

  • Reality over advocacy
  • Evidence over sentiment
  • Systems over noise

Internet governance is too often described as theory, consensus, or institutional comfort while operators live with the consequences of structure, scarcity, jurisdiction, and incentives.

BTW shows how the system works when assumptions are tested: clear, uncomfortable, evidence-led reporting that keeps attention on structure rather than drama.

Core Architecture

Governance, market, and membership branches operating in one integrated intelligence model.

Governance

  1. The Waiting List and the Secondary Market Are One System
  2. APNIC reclamation without surprise confiscation
  3. What APNIC's final IPv4 queue makes networks do while they wait
  4. Bankruptcy Remote, Registry Dependent
  5. IRP Awards and the Cost of Making ICANN Listen
  6. Virginia Corporate Law and the Reach of ARIN’s Registry Decisions
  7. AI Data Centres and the Return of Address Demand
  8. Inter-RIR Compatibility as a Private Trade Barrier
  9. Liquidity Without a Central Exchange
  10. Data Accuracy SLAs the RIRs Do Not Publish
View Governance

Market

  1. Granite Telecommunications and AS16504: Where Managed Network Control Ends and Business Continuity Begins
  2. Amazon Data Services Uruguay Is in the LACNIC Ledger, but AWS Outposts Is Not a Uruguay Cloud Region
  3. Inside The Internet, Inc and AS14473 show two announced IPv4 prefixes with uneven RPKI coverage
  4. Hispamar's Brazil Control Centre Shows Why Satellite Links Still Need Local Operating Proof
  5. Fiber57's October routing silence leaves Pimentel's recovery path unproven
  6. STPHNET Software Technology Park and the Evidence of a Quiet Network
  7. WHITETELECOM originates 36 IPv4 routes through one visible upstream, while Bonao's storm-ready last mile remains undisclosed
  8. Storm Iraq and the 333-Path Test: Where an Operator’s Control Really Ends
  9. Outofbox Cloud promises eight data-centre regions; public evidence still points to one Belagavi footprint
  10. Heartbleed made shared cryptographic maintenance a systemic-dependency accountability test
View Market

Membership

  1. Hostinger Operations UAB as Registrar: Can a Business Recover Its Domain When the Administrator Disappears?
  2. Linode and AS63949: Cloud Convenience Still Needs Operator Discipline
  3. SiteGround Spain, RIPE Membership and Centralized DNS: Who Controls Website Continuity?
  4. IONOS SE in the RIPE Ledger: Can Reverse DNS Survive an IP Address Handover?
  5. DreamHost and AS26347: Where Shared-Hosting Reliability Depends on DNS, Backups and Customer Control
  6. Liquid Web B.V. in the RIPE Ledger: Can an Abuse Report Reach the Right Operator?
  7. Opticomm's delayed service-area notices showed why a broadband build needs a live public handover record
  8. Telstra's 2020 local-number porting shutdown showed that keeping a number depends on an executable handover
  9. Ray Bellis and the DNS Work That Turns Silence into a Usable Answer
  10. Mohamed Nasr and the Route-Diversity Decisions Behind Egypt's Subsea Crossing Hub
View Membership

Number Resource Society

Reporting on address stewardship, representation, and the institutions shaping number-resource governance.

View Number Resource Society
  1. One Resource Holder, One Verifiable Control Claim
  2. When One Corporate Group Casts Twenty Registry Votes
  3. NRS Growth Without Mandate Laundering
  4. The Public Evidence Pack NRS Asks Registries to Publish
  5. A Global South Label Cannot Substitute for a Mandate
  6. Why NRS Advocates Records for Leased and Delegated Use
  7. Registry Policy by Demonstrated Operational Effect
  8. A Future Registry Worth Leaving
  9. Registry Resolution Before Insolvency: The Safeguards NRS Advocates
  10. Registry Service Levels Written for Customers, Not Dashboards

RIR Watchdog

Five regional sessions: ARIN, RIPE NCC, APNIC, AFRINIC, and LACNIC.

Network Operator Group (NOGs)

Operational intelligence from the communities where routing, peering, resilience, and deployment practice are tested.

View Network Operator Groups

History of the Internet

The internet's hidden roots in military strategy

Structural context for current governance behavior and infrastructure control models.

Photorealistic editorial risk and accountability image for Dyn made managed DNS a revenue-continuity accountability boundary

Dyn made managed DNS a revenue-continuity accountability boundary

The 2016 Dyn attack did not take down the internet. It did something narrower and more costly for dependent services: it made revenue, login, payment, media, support, and public-information flows unreachable because a shared authoritative DNS provider was overwhelmed. The accountability question is who controlled DNS diversity, transit exposure, customer continuity, and practical revenue-cycle workarounds before a managed-DNS dependency became a common-mode failure.

  1. DigiNotar made certificate-authority failure an operational-harm control test
  2. Operator Rights Outside the Multistakeholder Theatre
  3. Let's Encrypt made root expiration a certificate-trust accountability boundary
Open History Session

Leadership Alliance

Global AI summit launches as world enters implementation decade

Capital, policy, and institutional leadership alignment in cross-border digital systems.

Hostinger Operations UAB as Registrar: Can a Business Recover Its Domain When the Administrator Disappears?

Hostinger Operations UAB appears in both the RIPE NCC member directory and ICANN's accredited-registrar directory. Those records identify a company and a formal role. They do not answer the question that matters on a difficult Monday morning: can the business still control, renew, transfer and securely delegate its domain if the person who normally operates the account is gone?

  1. Linode and AS63949: Cloud Convenience Still Needs Operator Discipline
  2. SiteGround Spain, RIPE Membership and Centralized DNS: Who Controls Website Continuity?
  3. IONOS SE in the RIPE Ledger: Can Reverse DNS Survive an IP Address Handover?
Open Alliance Session