Iran’s IPv6 Blackout Underscores the Limits of Dual‑Stack Resilience

Iran has entered an IPv6 blackout, with major networks cutting off v6 reachability and forcing traffic back to IPv4. Under the hood, most consumers will still get online thanks to Happy Eyeballs and dual-stack fallbacks, but the change isn’t benign: IPv6-only endpoints go dark, latency can rise as failed v6 attempts time out, and routing/anycast behaviors shift as CDNs and resolvers adapt. What’s notable here is how a single toggle at the address-family level can reshape traffic patterns across an entire market without a full “internet shutdown.”

The bigger picture: this is a stress test for operators and developers who’ve treated IPv6 as “available enough.” It highlights operational and policy reality-IPv6 adoption is uneven, and in some jurisdictions it can be selectively withdrawn. Worth noting: public telemetry (think Cloudflare Radar, Google’s IPv6 adoption charts, RIPE Atlas) should show the cliff-drop in v6 while v4 stays flat, a telltale of intentional filtering or route withdrawal versus physical outage. For builders, the immediate takeaways are pragmatic: avoid IPv6-only control planes, keep dual-stack for critical services, tune connection racing to minimize failover drag, and ensure monitoring distinguishes v4/v6 health. Industry-wide, the event is a reminder that IPv6’s momentum doesn’t equal inevitability-operational and regulatory levers still matter.