17 Hormuz Cables, 30% of Global Internet: Names, Routes, Developer Failover

The internet feels weightless — cloud infrastructure, wireless signals, satellite connectivity. But 97% of all intercontinental internet traffic, and approximately 80% of US military communications, travel through physical cables lying on the ocean floor. Those cables are fragile, slow to repair, and increasingly in the crosshairs of geopolitical conflict.

The Middle East conflict of 2026 has brought renewed attention to a set of undersea cables that pass through one of the most strategically contested bodies of water on earth: the Strait of Hormuz.

The Cables That Matter

The Persian Gulf and surrounding waters host several critical undersea cable routes:

EPEG (Europe-Persia Express Gateway): A 25,000 km cable connecting Europe to the Middle East and onward to India and Southeast Asia. Passes through the Persian Gulf.

OMRAN: A 600 km cable connecting Oman to Iran. One of Iran's primary connections to the global internet.

UAE-Iran cables: Several shorter cables connecting UAE exchange points to Iranian landing stations — part of the routing infrastructure that was disrupted in the February 2026 attacks.

SMW5 (SeaMeWe-5): One of the world's longest submarine cable systems, connecting Singapore to France via the Middle East. Any disruption in Gulf waters affects this route.

To understand why these matter, consider where internet traffic actually flows. When someone in Europe calls up a website hosted in Singapore, that request almost certainly travels through undersea cables — and several of the shortest, lowest-latency routes pass through the Persian Gulf.

The Red Sea Precedent

In 2024, Houthi attacks in the Red Sea — a neighboring maritime chokepoint — severed three major undersea cables: AAE-1, Seacom, and EIG. The repair took months. During that period, internet traffic between Europe and Asia was rerouted through alternative paths, causing measurable latency increases and capacity constraints for users in East Africa, the Middle East, and South Asia.

If a similar disruption happened in the Strait of Hormuz — whether through deliberate attack, collateral damage from naval operations, or anchors dragged by vessels during conflict — the impact would be significantly larger. The Strait is narrower, the cable concentration is higher, and the geopolitical context makes rapid repair operations far more dangerous.

What "Repair" Actually Means

Undersea cable repair is not fast. A cable ship (a specialized vessel that can lay and splice cables at depth) typically takes 2-4 weeks to reach a break site, assess damage, and execute a repair — assuming the break site is accessible and politically safe to work in.

During active conflict:

• Cable ships may be unable to operate in the conflict zone
• Insurance for repair operations in war zones is unavailable or prohibitively expensive
• Military operations may restrict maritime traffic in the affected area
• Competing territorial claims may complicate which country's permission is needed to operate

The 2024 Red Sea repairs took roughly 2 months from initial break to full restoration. A similar break in the Strait of Hormuz during active hostilities could take significantly longer.

The Baltic Sea Pattern

The Persian Gulf isn't the only region where undersea cables are under threat. In late 2024 and into 2025, multiple Baltic Sea cables were mysteriously severed — including C-Lion1 (connecting Finland to Germany) and BSC East-West (connecting Latvia to Sweden). Investigations pointed to deliberate sabotage, though attribution remained contested.

Russia's naval doctrine explicitly treats undersea cables as legitimate targets in conflict. The pattern of "accidental" anchor drags by vessels near critical cable routes has repeated enough times to be statistically implausible.

This is a global pattern, not a regional one. The internet's physical layer is being recognized as a strategic vulnerability by every major military power.

What This Means for Developers Building Real Systems

Most developers never think about undersea cables. You don't need to — until the day your application's external API calls start timing out and your cloud provider's status page shows a partial outage in a specific region, and the root cause is a cable 3,000 meters below the surface of the Persian Gulf.

1. Understand your application's geographic dependencies

Most cloud services publish which undersea cable routes they rely on for cross-region data transfer. AWS, GCP, and Azure all have regional internet exchange points and published network path information. Know which cable routes your critical traffic relies on.

2. Multi-region architecture is not just about uptime — it's about cable paths

Running your database in us-east-1 and your application server in eu-west-1 means your traffic crosses the North Atlantic. Running both in ap-southeast-1 and eu-west-1 means Persian Gulf or Indian Ocean cables. Geographic redundancy should account for cable route diversity, not just data center diversity.

3. CDN configuration matters more than you think

A well-configured CDN routes requests through the path with the lowest latency and highest availability — but that path is determined by BGP and physical cable availability. When a cable goes down, CDN providers reroute automatically, but with higher latency. Your performance budgets should account for worst-case routing scenarios.

4. Third-party APIs are unprotected

If your application calls a payment gateway, authentication service, or data API hosted in a different region, that traffic crosses cables you don't control and can't monitor. Build timeouts and fallback behaviors for critical external dependencies. Assume any third-party API can become unavailable for days, not hours.

5. Consider latency budgets for conflict-adjacent regions

If you serve users in the Middle East, South Asia, or East Africa, their experience is particularly sensitive to Persian Gulf cable health. Consider regional CDN PoPs and edge caching specifically for these user segments.

The Subsea Cold War

Behind the immediate conflict, there's a longer-term strategic competition underway. The US, UK, China, and Russia are all actively mapping, monitoring, and — in some cases — physically accessing undersea cable infrastructure.

The US and UK operate specialized naval vessels capable of tapping undersea cables. China has been expanding its presence in cable landing station ownership and the cable manufacturing supply chain. Russia's submarine activity near cable routes has been documented by NATO.

This isn't conspiracy theory — it's documented in military doctrine, congressional testimony, and academic research on undersea infrastructure security. The cables that carry your application's traffic are simultaneously critical infrastructure, intelligence collection assets, and military targets.

The Practical Bottom Line

You can't protect the cables. You can build systems that degrade gracefully when they fail.

The patterns that work: multi-region architectures with cable-diverse routing, aggressive edge caching to reduce cross-ocean requests, meaningful timeout and fallback logic for external dependencies, and monitoring that distinguishes between application errors and infrastructure-level connectivity issues.

The conflict in the Middle East is a reminder that the physical internet has physical vulnerabilities. Engineering for that reality is not paranoia — it's good architecture.

Related: Iran's Internet Collapsed to 4% of Normal — Technical Breakdown | USA-Israel Strikes on Iran: Tech and Cyberwar Impact