Cloudflare records global failures and affects access to sites like X and Spotify this Tuesday (18)

The company Cloudflare, a provider of security and content distribution services on the internet, is facing technical instability this Tuesday, November 18, 2025. The problem, which started around 12:00 UTC, affects multiple customers on a global scale, with reports of 500 errors on websites and applications. Users in Brazil and other countries report difficulties accessing platforms dependent on the company’s network.

The peak of complaints occurred between 12pm and 1pm, Brasília time, according to outage monitoring data. Services such as X (formerly Twitter), Spotify and ChatGPT are slow or completely unavailable in several regions. Cloudflare confirmed the investigation into the incident, which also compromised the company’s own dashboard and API.

These errors arise during scheduled maintenance at the Santiago, Chile datacenter, scheduled between 12pm and 3pm UTC. Experts point to a possible relationship between the routine and the failure, although the company does not detail the initial cause. The impact affects around 68% of complaints related to connections to servers.

User reports grow in a few hours

Users began noticing the glitches shortly after noon, with messages about pages not loading. Outage tracking platforms, such as DownDetector, have recorded spikes above the normbadly, with more than 2 thousand notifications in Brazil.

The outage affects not only end consumers, but also companies that use Cloudflare’s reverse proxy to protect their websites. Social networks and productivity tools lead the complaints.

Services impacted in detail

Several websites and applications depend on Cloudflare infrastructure for daily operation. The widespread failure compromises data traffic between visitors and hosts.

• X (Twitter): Users are unable to update timelines or post content in multiple regions.
• Spotify: Playing music and podcasts has loading errors.
• OpenAI (ChatGPT): Access to web interface fails for real-time queries.
• League of Legends: Multiplayer game servers record disconnections.
• Canva: Online design tool crashes during collaborative edits.

These services represent a fraction of the thousands affected, as Cloudflare serves 20% of global web traffic.

Technical investigation progresses with updates

Cloudflare issued initial statements at 12:03 UTC stating that it is reviewing the full scope of the issue. Teams work with third-party providers to mitigate 500 errors common in backend failures.

UpdatesSubsequent updates, around 12:17 UTC, highlight that responses to customer queries remain operational via live chat. The company plans email notifications and webhooks for paid plans.

Previous maintenance, such as the one in June 2025, lasted up to 2 hours and 28 minutes, affecting services such as Workers KV and WARP. This time, the focus is on restoring stability without prolonged outages.

Incident history reveals patterns

Similar failures occurred in 2025, such as the one on June 12, which took down APIs and dashboards for almost three hours. In that episode, the problem was rooted in a change in service topology.

Leia Tambem

Michael Schumacher: A Decade of Private Recovery and an Enduring Legacy in May 2026

Cometa interestelar 3I/Atlas revela segredos cósmicos em sua passagem próxima à Terra

Sony reduz preço do controle DualSense Edge para expandir base de jogadores no PlayStation 5

In August, traffic congestion for customers on AWS us-east-1 caused severe slowdowns. These events highlight the internet’s dependence on few core infrastructures.

Quarterly reports from Cloudflare indicate that global outages account for less than 1% of annual uptime, but each one affects millions of hits.

Contingency measures for users

Affected companies activate backup plans to minimize losses. Individual users can test connections via alternative VPNs or browsers in incognito mode.

Cloudflare recommends monitoring the official status page for real-time updates. Diagnostic tools like DNS tests help identify whether the problem is local or global.

Partial resolution restores access startsis

At around 13:00 UTC, reports indicate recovery in some regions of Europe and Asia. In Brazil, traffic to X and Spotify gradually improves, with 40% fewer complaints registered.

The company confirms that the dashboard returns to normal for business accounts. Investigations continue to prevent recurrences, focusing on network redundancies.

Engineers adjust configurations in affected datacenters, prioritizing critical traffic. The incident reinforces the need for diversification in CDN providers.

Prevention in focus for the sector

Software updates and load testing are part of Cloudflare’s routines to avoid failures. Partnerships with third parties guarantee 24-hour incident support.

Corporate users access dedicated emergency lines. Transparency in communications helps manage expectations during crises.