Users from different countries faced difficulties accessing digital platforms this Tuesday morning (18). Sites like the social network X, OpenAI’s ChatGPT and the Amazon store showed simultaneous instability. The problem also affected services such as Canva and Discord, according to registered complaints.
Cloudflare, a company responsible for managing traffic and protecting millions of websites, confirmed that it is investigating a flaw in its infrastructure. The outage directly affected the functioning of platforms that use its services. The peak of complaints occurred around 8:30 am Brasília time.
- Main affected platforms:
- Social network X (formerly Twitter)
- ChatGPT and OpenAI services
- Amazon and Amazon Web Services
- Canva, Discord and Grammarly
- Banking and retail websites in various countries
Cloudflare admits immediate investigation
The company published an update on its status page indicating that engineers are working to identify the cause. The issue involved partial interruption in the CDN service and protection against DDoS attacks.
The failure lasted around 40 minutes at the most critical time. Usuários reported error 500 and timeout messages when trying to open the affected platforms.
Complaints soar at Downdetector
The Downdetector website recorded more than 100,000 notifications for the X network alone in the early hours of the morning. Amazon received around 40 thousand complaints in the same period.
ChatGPT had a peak of 30 thousand reports of instability. Downdetector itself experienced momentary slowdowns during the incident.
Problems were reported in América of Norte, Europa and América of Sul simultaneously.
What Cloudflare does on a daily basis
The company acts as an intermediary between servers and end users across 24 million domains. Seu system distributes content through data centers in more than 300 cities.
The technology protects against cyber attacks and speeds up page loading. Grandes platforms depend on the service to maintain constant availability.
History of similar incidents
Cloudflare has already faced global outages in 2022 and 2023 that affected multiple services. The June 2022 incident lasted about an hour and a half.
Previous failures involved routing and network configuration issues. The company implemented improvements after each occurrence.
Gradual return of services
The X network returned to normal operation around 10 am Brasília time. ChatGPT reestablished full access minutes later.
Amazon reported that its core services are operating without interruption at this time. Cloudflare maintains active monitoring to prevent new episodes.
Increasing dependence on single infrastructure
Millions of companies use a few cloud providers around the world. Concentration increases the impact when failure occurs at a single point.
Experts point out that the current internet architecture facilitates the rapid spread of technical problems. Diversificação from suppliers appears as an alternative discussed in the sector.