The Longest Route We've Ever Measured: Oman to Chile at 452ms

Based on real RIPE Atlas measurements from GeoCables monitoring infrastructure, March–April 2026.

At 452 milliseconds average, the route from Oman to Chile is the highest-latency path in our entire measurement database. But what makes it truly remarkable isn't the number — it's the journey. A packet leaving Barka, Oman doesn't take the shortest path across the Indian Ocean. It travels east to Singapore, crosses the Pacific to Japan, cuts across the continental United States, and only then crosses into South America to reach Valparaíso, Chile.

Total distance traveled: approximately 35,000 km. The circumference of Earth is 40,075 km. The actual end-to-end great-circle distance between Barka and Valparaíso is about 16,800 km — the shortest path a packet could theoretically take. The route we measure is more than twice that.

In an idealised world where the global internet honoured geography, a packet from Oman to Chile would hop south across the Indian Ocean, cross the Southern Ocean basin, and emerge in South America via Australia or directly via the Antarctic-fringe cables that have been proposed but never built. In the actual internet of 2026, no such route exists, no major carrier offers it, and BGP — the protocol that selects paths — does not even consider it as an option. The packet goes around the long way because that is the only way the routing layer knows.

This article walks through that journey: the measurements that establish the latency, the traceroute that reveals the path, the cable systems involved, the BGP economics that produced this routing, and what an optimised version of the same trip would look like.

The Route in Numbers

Our RIPE Atlas probe in Barka, Oman has been measuring this route continuously since early March 2026, with measurements every ~15 hours:

The baseline is ~427ms, but on March 5–7 we observed a significant spike to 514–564ms — a 32% increase suggesting congestion or partial rerouting on one of the submarine cable segments. Six weeks later, in late April, the route had returned to baseline and stayed there. Whatever caused the early-March anomaly cleared up; the underlying path remained unchanged.

The Real Path: Traceroute Reveals Everything

A traceroute from our probe in Barka reveals the complete picture:

The most dramatic moment in this traceroute is hop 8: the RTT jumps from 79ms in Singapore to 409ms — a single leap of 330ms. This is where the packet enters a transpacific submarine cable system operated by TATA Communications, crossing from Southeast Asia to Japan, and from there onward across the Pacific to Los Angeles. That single submarine hop alone accounts for more time than the entire Oman-to-Singapore segment that preceded it.

Three Oceans in One Route

Leg 1 — Indian Ocean (Barka → Singapore):
Zain Omantel's network carries traffic from Muscat eastward. The RTT of 79ms for the Muscat–Singapore segment (~5,900 km) is consistent with the <a href="/cable/seamewe-5">SeaMeWe-5</a> or AAE-1 cable systems, both of which have landing points near Muscat and in Singapore. SeaMeWe-5 is a 20,000-km Asia-Europe trunk that lands at Barka itself; AAE-1 is the newer 25,000-km Asia-Africa-Europe-1 system. Either is consistent with the 79ms measurement to Singapore. Both cables are also on a path that has hosted notable connectivity issues in the past — the same general geography that produced the connectivity story we documented for Turkmenistan's two-gateway isolation, where landlocked Turkmen traffic transits through Azerbaijan and Uzbekistan to reach the same global backbone.

Leg 2 — Pacific Ocean (Singapore → Los Angeles):
TATA Communications (AS6453) takes over in Singapore. The massive RTT jump at hop 8 (79ms → 409ms) marks the entry onto a transpacific cable — most likely TGN-Pacific or a similar TATA-operated system. The packet surfaces in Tokyo before heading to Los Angeles, covering roughly 15,000 km of open ocean. This is the longest single submarine segment of the route and contributes the largest fraction of the total round-trip time. Pacific transits through Japan are a long-established TATA pattern; they exist in part because TATA inherited a substantial Asia-Pacific submarine portfolio from its acquisition of the former VSNL international assets, and routing through that portfolio is structurally cheaper for TATA than handing traffic to a competitor at Singapore.

Leg 3 — Americas (Los Angeles → Santiago → Valparaíso):
From Los Angeles, the traffic travels across the continental United States to Dallas, then to Miami, where it enters the Caribbean and South American network. From Miami, it crosses into Chile via a cable likely connecting to SAM-1 (South America-1) or <a href="/cable/south-american-crossing-sac">South American Crossing (SAC)</a>, both of which land in Valparaíso. Miami's role as the de-facto internet exchange for most South American traffic is a structural feature of the region's connectivity economics — almost every Latin American carrier maintains its primary peering relationship with Tier-1 networks in Miami, regardless of where its actual subscriber base is located.

Why Not the Shorter Path?

Option A — Westward through Europe and the Atlantic (~280ms theoretical):
Oman → Red Sea → Suez Canal → Mediterranean → Atlantic → Brazil → Chile. This would use cables like EIG, AAE-1, or SEA-ME-WE 4/5. Estimated RTT: 280–320ms. The path is geographically reasonable and uses well-engineered cables, but it requires a carrier handoff between East-of-Suez and Latin-American transit providers that does not currently have the commercial density to attract general internet traffic. The same kind of "geographically reasonable but commercially absent" pattern shows up across the Indian Ocean: the route from Azerbaijan to Kazakhstan via the Trans-Caspian seafloor is shorter than the actual measured paths but is not used because no carrier has built the peering relationships to make it cheap.

Option B — Direct Indian Ocean and Southern Pacific (~200ms theoretical):
Barka is connected to the Oman-Australia Cable (OAC), which opened in 2022. From Australia, cables like <a href="/cable/hawaiki">Hawaiki</a> connect to New Zealand and onward to Chile. Valparaíso has landing points for SAM-1, SPCS<a href="/cable/mist">MIST</a>RAL, and SAC. The direct path exists — but BGP ignores it.

Why BGP Routing Ignores Geography

1. TATA's global backbone dominance:
Once Zain Omantel hands off traffic to China Mobile in Singapore, and China Mobile hands it to TATA, the packet follows TATA's internal routing — which sends South American destinations via the Pacific to the US, then down through the Americas. This is not a routing failure; it is a routing decision based on TATA's existing peering and transit relationships. TATA terminates traffic in Miami extremely efficiently because Miami is where TATA's South American peers expect to receive traffic. Sending the packet via Australia and the Southern Pacific would require a carrier-to-carrier handoff that does not currently exist for this kind of general internet flow.

2. No direct peering to OAC:
The Oman-Australia Cable is relatively new and primarily serves Oman–Australia enterprise traffic. It doesn't yet have the carrier relationships to route general internet traffic from Oman to South America. This is the chicken-and-egg problem of new submarine infrastructure: until a cable has dense peering, traffic doesn't use it; until traffic uses it, peering doesn't develop.

3. Traffic aggregation in the US:
Miami is the de facto internet hub for South America. Major carriers route Latin American traffic through Miami regardless of origin. Roughly 80% of Latin American international traffic passes through Florida, according to TeleGeography's regional capacity reports. As long as that gravitational centre exists, paths that bypass it remain commercially fragile and hence rarely chosen by BGP.

TATA, Inheritance, and Backbone Politics

The TATA backbone deserves a closer look. AS6453 — the autonomous system that carries the majority of our measured Oman-Chile route — is the international transit arm of Tata Communications, formed in part from the assets of VSNL (Videsh Sanchar Nigam Limited), the former Indian state-owned international carrier privatised in 2002. When VSNL was acquired by Tata, it inherited a globally significant submarine cable portfolio: stakes in transpacific systems, transatlantic systems, and Asian regional cables. That inheritance is why TATA today is the natural transit choice for Asia-to-Americas traffic that crosses the Pacific.

The economic logic for TATA is straightforward. Each kilometre of submarine fibre that TATA already owns or has IRU rights on is paid-for capital. Routing traffic across that capital is cheap. Handing traffic to a competing carrier, by contrast, requires settling a transit invoice. Given a choice between routing Oman→Chile via Singapore→Japan→LA→Miami→Santiago (all on TATA's own infrastructure) or via a multi-carrier path through Australia and the South Pacific, TATA always chooses the on-net path. The 452ms is the cost of that choice rendered as latency.

Oman's Submarine Connectivity Picture

Oman is not poorly connected. The country has multiple submarine cable landings, including SeaMeWe-5, AAE-1, the Oman Australia Cable (OAC), the proposed BICS Africa-2, and several regional Gulf systems connecting to the UAE and to Iran. Total inbound capacity comfortably exceeds 50 Tbps. The issue is not capacity. The issue is which capacity gets used for which destination.

For traffic to Chile, none of Oman's cables route directly. SeaMeWe-5 carries Oman to Asia and Europe, not to South America. AAE-1 carries Oman to East Asia and to Marseille, not to South America. OAC carries Oman to Perth, but no major carrier offers Perth-to-Santiago transit at meaningful scale. Each of Oman's submarine cables is solving a different connectivity problem; none of them is solving Oman-to-Chile.

The March 5–7 Anomaly

The spike from 426ms to 564ms over March 5–7 — a 138ms increase — suggests congestion on the TGN-Pacific segment, partial rerouting due to a cable fault, or scheduled maintenance pushing traffic to backup paths. Recovery to baseline by March 8 confirms temporary congestion rather than a cable break. The fact that April measurements stabilised at the same baseline confirms that whatever happened in early March was transient. The route's 452ms steady-state appears to be its 2026 baseline, not a perturbation from a brief incident.

What an Optimal Route Would Look Like

Oman → OAC → Australia → Southern Cross / Hawaiki → Chile:

• Barka → Perth via OAC: ~65ms

• Perth → <a href="/location/sydney-nsw-australia">Sydney</a> → Auckland → Valparaíso: ~140ms

• Total theoretical RTT: ~200–220ms — a 2× improvement over current 427ms.

For this route to actually exist as a commercial product, several things would need to happen: a Tier-1 carrier would need to extend its presence to OAC's Perth landing; that carrier would need direct peering with Australian backbone networks; and onward transit from Australia to Chile would need to develop the kind of commercial density that currently exists only for Australia-to-US-West-Coast. None of this is technically difficult; all of it depends on commercial volume that currently does not exist on this corridor.

Monitoring Status

GeoCables monitors the Barka → Valparaíso route via RIPE Atlas probe 65614, every ~15 hours.

• Current RTT: 427–433ms (April 2026 baseline) | Best: 423ms | Peak: 564ms (anomaly Mar 7)

• Path: Zain Omantel → China Mobile → TATA Communications → Telmex Chile

This route remains the highest-latency path in our database — proof that in the global internet, geography and routing are two very different things. The packet does not travel the path a map would suggest. It travels the path that twenty years of commercial backbone evolution have laid down between the carriers involved. The 452 milliseconds is not a measurement of distance. It is a measurement of how that history routes around the geography it inherited.