Cloudflare, an internet infrastructure provider, recorded a global outage in its control panel and APIs this Friday, December 5, 2025. The problem began around 8am UTC, equivalent to 5am in Brasília time, and affected user requests for dashboards and software integrations. Plataformas dependents, such as Zoom and Shopify, reported access errors during the peak of the failure.
Monitoring reports indicate that internal degradation impacted around 20% of global web traffic managed by the company. The second occurrence in less than three weeks, following a similar incident on November 18, generated error 500 messages across multiple services. Especialistas attribute the event to conflicts in scheduled maintenance in data centers.
The incident occurred during peak times in regions such as Europa and Ásia, where morning traffic coincided with the start of the failure. Usuários companies faced difficulties with management tools, while ordinary consumers noticed slowness on e-commerce and collaboration sites.
- Key affected services include Zoom, for video calls;
- Shopify, with online checkout failures;
- Notion and Medium, displaying page loading errors.
Details of the technical failure
Engineers at Cloudflare identified the issue as an internal service degradation that began at 8:56 UTC in data centers at Chicago and Detroit. Manutenções scheduled in these locations, between 7am and 11am UTC, collided with a glitch in the scripts of Workers, a task automation tool. The company deployed partial fixes at 9:12 am UTC, restoring 70% of operations within an hour.
Monitoring platforms, such as Downdetector itself, were temporarily inaccessible due to dependence on Cloudflare. Error 500, which signals server failures, has spread to API integrations used by developers around the world. Até at 10am UTC, user reports on América and Norte indicated recovery in 80% of cases, but Asian regions were still experiencing delays.
The flaw highlighted the interconnection of web services, where an outage at edge layers affects millions of domains simultaneously. Cloudflare manages traffic for 20% of global websites, including DDoS protections and content acceleration.
Services impacted in depth
Zoom has confirmed outages in logins and active sessions, with users on Europa reporting outages at 9am CET. The platform, which relies on Cloudflare APIs for media distribution, saw 40% spikes in errors in video calls during the incident.
E-commerce-focused Shopify has faced blockages on checkout pages, potentially affecting transactions at peak times in the US. Relatórios internals indicate that 15% of online stores stopped processing orders for 45 minutes.
Canva and Notion, design and productivity tools, displayed error messages in uploads and collaborative edits. In the case of Notion, database integrations failed, impacting remote teams at Ásia and Oceania.
Medium, a publishing platform, reported slow article loading, with users on Índia reporting complete unavailability at 2:30 pm IST.
Recent outage history
Cloudflare experienced a similar outage on November 18, 2025, when a bug in bot management files caused four hours of unavailability. Naquele event, services such as X and ChatGPT went offline for 90 minutes, affecting 10 million global hits.
In September 2025, another technical failure resulted in traffic slowing down to 5%, linked to overload on edge servers. Esses incidents accumulate in a year marked by growing traffic demands, with global volume rising 25% since 2024.
The company revised configuration protocols after November, implementing automated tests for dynamically generated files. However, the recurrence points to challenges in scalability during parallel maintenance.
Consulting reports indicate that interruptions like this cost an average of US$9,000 per minute for business clients, due to losses in productivity and sales.
Mitigation measures adopted
Cloudflare technicians activated failover routes in secondary data centers on Europa and Ásia, redirecting 60% of affected traffic within 20 minutes. Scripts from Workers were isolated to prevent propagation, with real-time monitoring via the status API.
Customers received alerts via dashboard at 9am UTC, advising the use of local caches to minimize impacts. The company expanded capacity by 15% at nodes Detroit and Chicago, concluding maintenance at 1pm UTC with no further failures.
Developers were advised to implement retries in API calls, reducing errors in third-party integrations. Plataformas and Zoom restored full services at 10:30 UTC, confirming stability in global tests.
Prospects for future stability
Software updates planned for December include double validations on bot configurations, aiming to eliminate similar bugs. Cloudflare invested US$500 million in network expansion in 2025, adding 50 new data centers.
Infrastructure experts recommend provider diversification to mitigate risks, although dependence on Cloudflare persists in 80% of large platforms. Continuous Monitoramento via tools like PagerDuty now integrates webhooks for proactive alerts.
Full restoration occurred at 11am UTC, with traffic returning to normal across 95% of domains. The company will publish a detailed report on its official blog within the next 24 hours, detailing lessons learned.
Reactions from users and companies
Business users have expressed concerns on developer forums about recurrence of crashes. Equipes of IT in e-commerce reported losses estimated at US$ 2 million during the peak.
Affected platforms such as Shopify have issued internal advisories to vendors suggesting offline backups. In the collaboration sector, Notion has prioritized status updates for premium subscribers.
The cybersecurity community has highlighted the irony, given the recent blocking of a 29.7 Tbps DDoS by Cloudflare. Discussões in professional networks emphasize the need for redundancy in critical infrastructures.
Global network expansion
Cloudflare operates in 310 cities, covering 95% of the world’s connected population. In 2025, processed traffic reached 100 exabytes per month, an increase of 30% compared to 2024.
Investments in edge computing enable local processing, reducing latency by 40% for users in the Ásia-Pacific. Parcerias with local operators in Bogotá and Santiago expanded coverage in América Latina.
The company maintains compliance with regulations such as GDPR and LGPD, ensuring privacy in transited data. Atualizações recent edge encryption protects 99.99% of connections from interception.
