Inglês News

Outage in Cloudflare infrastructure interrupts access to Metropoles, NSC Total, PortalLeoDias and other sites around the world

Cloudfalre leo dias

A technical instability at Cloudflare, an internet infrastructure provider, interrupted access to several websites around the world at around 11am this Tue

Cloudfalre leo dias - Internet

A technical instability at Cloudflare, an internet infrastructure provider, interrupted access to several websites around the world at around 11am this Tuesday, November 18, 2025. The company confirmed the issue on its status dashboard, attributing it to an internal degradation in the global service, without details on the initial cause.

The incident affected not only media outlets, but also social networks and essential services that depend on Cloudflare’s content distribution network. Company’s Técnicos began immediate investigations, with updates indicating gradual recovery in some regions. Usuários on Brasil reported 500 errors and connection failure messages, impacting daily information traffic.

Cloudfare - Downdetector
Cloudfare – Downdetector

Services affected on Brasil

News portals have faced direct outages due to their reliance on Cloudflare for traffic routing and protection. Metropoles, for example, displayed blank pages for thousands of simultaneous accesses.

NSC Total, which focuses on regional content from Santa Catarina, also stopped responding, forcing writers to move to alternative channels temporarily. Portal Léo Dias, known for exclusive celebrity coverage, has seen spikes in complaints on parallel networks.

Other outlets, such as UOL and G1, operated with partial instability, but maintained printed editions and live broadcasts intact.

Cloudflare Technical Responses

The company issued short statements confirming the impact on services such as Workers and CDN. Engenheiros identified the issue as a global network failure and implemented fixes in key data centers.

Subsequent updates reported normalization in hits for WARP and Access, but error rates persisted in APIs. Scheduled maintenance on Santiago, on Chile, coincided with the spike, although not confirmed as a direct cause.

Corporate customers received alerts via dashboard, prioritizing restoration of critical traffic.

  • Widespread errors 500 affected data routing;
  • Cloudflare dashboard was inaccessible for 20 minutes;
  • Third-party APIs, such as payment APIs, have experienced delays of up to 30 seconds.

Impacts on international platforms

X, formerly Twitter, displayed unloaded content to global users, including on Brasil. Plataformas like ChatGPT and Canva stopped responding, with error messages on login attempts.

Cloud services like AWS and Azure have faced ripple effects, disrupting integrations with digital banks. No financial sector, transactions via Nubank and Bradesco apps recorded occasional failures.

  • X: Lentidão in feeds and posts for 40 minutes;
  • ChatGPT: total Bloqueio in AI queries;
  • Canva: Ferramentas of inaccessible design in collaborative editions.

History of similar failures

Previous incidents highlight vulnerabilities in centralized infrastructures. In August 2025, a customer traffic saturation isolated connections to AWS, affecting thousands of websites.

Failures in 2024 involved software updates that propagated errors to partner networks. Esses events reinforce the need for diversification in CDN providers.

Publicly-listed Cloudflare reported rising profits in 2025, but cuts to engineering teams coincided with spikes in instability.

Experts note that 16 million users in 60 countries were impacted in past outages, similar to today.

Preventative measures for users

Companies recommend monitoring via independent network status tools. Usuários can switch to stable VPN connections during failures.

  • Check status on sites like DownDetector, when accessible;
  • Use browsers with offline caching for essential content;
  • Contact internet providers for alternative routes.

Full recovery occurred in stages, with 80% of services normalized by noon. Técnicos of Cloudflare continued adjustments in remaining regions.

User reports on Brasil

Residents of São Paulo and Rio of Janeiro shared experiences of interruptions in morning access. Um journalist from Metropoles reported migration to emails to send urgent matters.

In Florianópolis, NSC Total readers accessed issues via mobile apps, which used secondary servers. The Portal Léo Dias saw an increase in searches for updates from competitors.

These testimonies illustrate the daily dependence on digital infrastructures for professional and informational routines.

To Top