Date of incident: June 21, 2022
Affected entity: Cloudflare (global CDN, DNS, and edge-security platform)
Summary:
On June 21, 2022, Cloudflare experienced a major global outage that disrupted access to a wide range of websites, APIs, and online services. The failure resulted from a misconfiguration during a planned network update to Cloudflare’s edge routing infrastructure. A configuration change — intended to increase network resilience — instead triggered cascading failures across multiple critical data centers, causing widespread service interruption. Cloudflare restored full service approximately one hour after the issue began.
What Happened
During a routine maintenance operation, Cloudflare deployed a network configuration change to a portion of its backbone. This change affected the “spine” of Cloudflare’s system — the super-core data centers responsible for global routing. The new configuration consumed more resources than anticipated, overwhelming certain routers and producing a rapid feedback loop of instability.
The problem propagated quickly across Cloudflare’s backbone region due to the interconnected nature of its global routing layer, resulting in immediate global impact.
Impact
- Numerous websites and services relying on Cloudflare for DNS, CDN, and security protection became unreachable.
- Users saw “500 errors,” timeouts, and routing failures across large regions.
- API services, SaaS platforms, and enterprise applications behind Cloudflare experienced significant downtime.
- The outage lasted approximately 69 minutes before full stability was restored.
Because Cloudflare sits in front of millions of domains, even a short disruption produced noticeable global ripple effects.
Cause of the Outage
The root cause was a misconfigured network change to Cloudflare’s backbone data centers — specifically within its internal BGP architecture (“AS13335”). The rollout did not proceed through a phased, canary-style deployment. As a result, the configuration error was applied too broadly, too quickly, leading to immediate systemic instability.
Response
Cloudflare engineers:
- Detected the issue rapidly within internal telemetry.
- Performed urgent rollback actions to revert the routing configuration.
- Stabilized spine routers in affected regions.
- Conducted a detailed post-incident review that publicly documented the cause, timeline, and corrective actions.
Cloudflare committed to:
- Strengthening change-management controls
- Introducing stricter staged rollouts
- Improving the safety margins in routing configuration validation
- Enhancing internal monitoring that detects backbone instability even earlier
Significance
This incident demonstrates the operational fragility of global Internet infrastructure and the risks inherent in updating core routing systems. Because Cloudflare is one of the Internet’s largest edge networks, even an hour-long outage caused measurable disruption across critical services worldwide.
The event reinforced the importance of:
- Phased deployments
- Guardrails in automated configuration systems
- High-confidence rollback capabilities
- More conservative procedures for changes affecting backbone routers
Source(s)
Cloudflare. (2022, June 21). Cloudflare outage on June 21, 2022 — What happened?
https://blog.cloudflare.com/cloudflare-outage-on-june-21-2022/