Large parts of the internet experienced significant disruption today after Cloudflare, one of the world’s most widely used internet infrastructure providers, suffered a major outage. The issue has affected a wide range of services including X, Spotify, OpenAI, Canva, Facebook, LinkedIn... and even Downdetector, the platform normally used to track such outages, struggled to stay online.
Users across the UK and globally reported 500-error messages, failed requests and services that simply refused to load.
The issue began shortly after 11:20am UK time, when Cloudflare detected what it described as a "spike in unusual traffic" to one of its services.
This triggered an internal service degradation which quickly led to widespread errors across parts of Cloudflare’s global network.
Here are the key moments so far:
Cloudflare identifies a sudden increase in unusual traffic to one of its core services.
The company confirms reports of widespread 500 errors, dashboard failures and API timeouts.
Investigation continues. Error rates rise. The impact broadens.
Although partial recovery has been observed, issues persist and Cloudflare has warned of higher-than-normal error rates.
Downdetector becomes unstable, limiting visibility into outage reports as demand spikes.
Cloudflare temporarily disables WARP access in London as part of remediation work.
Cloudflare announces the issue has been identified and a fix is being deployed.
Access and WARP services are restored and WARP access is re-enabled in London.
Cloudflare continues restoring broader application services for affected customers.
Dashboard functionality is restored and application-level remediation continues.
Cloudflare states that a fix has been implemented and the incident is believed to be resolved, with monitoring underway.
While recovery is well underway, Cloudflare notes that some users may still notice intermittent issues as services continue to stabilise.
Cloudflare underpins a substantial portion of the modern internet.
Its services sit behind content delivery, DNS, DDoS protection, application security and the day-to-day operation of countless cloud-based platforms.
Because Cloudflare is embedded so deeply across the web, an outage of this scale is immediately visible, even to organisations that never interact with Cloudflare directly.
For businesses with 25 to 250 users, the events of today underline several practical realities.
1. Indirect dependencies can become immediate points of failure
Many organisations depend on tools that rely on Cloudflare behind the scenes. CRMs, HR systems, online portals, billing systems, authentication services and collaboration tools may all use Cloudflare’s infrastructure without the organisation ever realising it.
When an outage occurs, those hidden links become visible very quickly.
2. Service reliability is tied to customer perception
Clients may never know that Cloudflare is part of your service. What they notice is the result. If a platform slows down or becomes unavailable, trust and user experience are still affected, even when the root cause is outside your environment.
3. Cloud resilience is not guaranteed
Major outages across the industry have demonstrated the extent to which modern organisations depend on external providers. Rather than assuming uninterrupted uptime, planning for continuity is now essential.
Cloudflare is updating its incident status at cloudflarestatus.com throughout the day, where users can see which services have recovered and where issues remain.
Several key lessons stand out.
At Grant McGregor, we help organisations strengthen their resilience by:
Resilience becomes most visible when external systems fail and your organisation is still able to function.
Message us: https://www.grantmcgregor.co.uk/contact-us