A cluster of undersea cable breaks in the Red Sea slowed internet traffic across parts of the Middle East and South Asia this weekend, prompting Microsoft to reroute Azure workloads to maintain service. The company said Saturday, September 6, that customers whose traffic normally traverses the region could see higher latency, particularly on paths linking to Asia and Europe. By Saturday evening, Microsoft reported that it was no longer detecting active Azure incidents, though detours around damaged fiber can still add higher latency.
The individuals or group responsible for cutting the cables remains unknown. It’s possible that anchors dropped from ships can damage cables, though this was more likely a deliberate act. Network monitors and carriers corroborated widespread impact. NetBlocks flagged a series of subsea outages degrading connectivity in multiple countries, including India and Pakistan. Pakistan’s national operator warned of congestion during peak hours after breaks near the Saudi port of Jeddah. Other cloud and hosting providers also noted delays, including Linode.
What Enterprises Should Watch
Most organizations don’t need to scramble. As Microsoft noted in its official update (which is refreshed constantly) “We do expect higher latency on some traffic that previously traversed through the Middle East. Network traffic that does not traverse through the Middle East is not impacted.”
IT teams running latency-sensitive services or tight replication windows should review telemetry from Friday night through the weekend, verify failover policies, and scan provider advisories for any region-specific caveats. If a company’s network makes major use of a single Middle East transit route, even indirectly through a CDN or SaaS, it might have been impacted.
For Microsoft’s part, the company says it will keep rebalancing traffic while cable owners mobilize repair ships. Expect incremental improvements as routing settles and capacity returns. Repairing subsea infrastructure is challenging work. Faults must be pinpointed, cable ships dispatched, and splicing performed in weather windows. Repairs are typically finished in days or weeks, not hours.
The Red Sea has seen breaks before: incidents in February 2024 disrupted Asia–Europe traffic, and continual regional tensions increase the risk profile. Saturday’s statements did not attribute responsibility, and past claims of sabotage have been denied by groups accused of targeting cables. For now, operators are focused on regaining services rather than attribution.
Redundancy is Essential
If there’s a larger point here, it’s support for the long-established concept of cloud redundancy. Azure and, reportedly, most global services, stayed up by using the same network diversity that enterprises have typically created for themselves: multi-network transit, dynamic traffic engineering, and region-to-region failover. On the other hand, extra links add measurable delay to highly time-sensitive network operations like real-time collaboration and financial trading.
Cables on the seafloor carry the overwhelming majority of intercontinental data. When several are cut in the same corridor, the internet’s usual resilience is tested but can remain active: traffic gets pushed onto fewer remaining routes. However, this boosts jitter and round-trip times even when applications stay online. Microsoft, again, emphasized that workloads not crossing the Middle East were unaffected, and that it had shifted traffic to alternative paths while it optimizes routing.
