Outage History
A chronological record of significant internet infrastructure failures and service disruptions in 2024. Each incident includes root cause analysis, duration, and scope of impact.
ChatGPT and the OpenAI API experienced a multi-hour outage, causing 503 and 429 errors across all tiers. The outage was particularly disruptive given the high number of production applications reliant on the OpenAI API.
Discord users across North America and Europe reported inability to send messages, join voice channels, or load servers. The outage coincided with high-traffic periods and was resolved within a few hours.
A faulty content configuration update to the CrowdStrike Falcon sensor caused an estimated 8.5 million Windows machines worldwide to crash with a Blue Screen of Death. Airlines, hospitals, banks, broadcasters, and emergency services were among the most severely affected. The incident was widely described as the largest IT outage in history.
A disruption to Cloudflare's R2 object storage service caused widespread issues for websites and applications that depend on it for assets and media. Cloudflare published a detailed post-mortem attributing the issue to a misconfigured network device.
An electrical event at an AWS data center in Northern Virginia caused disruptions to EC2 instances, RDS databases, and several managed services. Customers running workloads exclusively in us-east-1 without cross-region failover were most affected.
Facebook, Instagram, WhatsApp, Messenger, and Threads all went down simultaneously for around two hours. Hundreds of millions of users were unable to log in or use the services. Meta attributed the outage to a technical issue during a configuration change.
GitHub experienced significant degradation affecting Actions, repository clones, webhooks, and API availability. The incident impacted CI/CD pipelines worldwide, blocking deployments for developers and engineering teams.
Data sourced from public post-mortems, official status pages (statuspage.io, instatus.com), and credible tech reporting. Duration estimates are approximate based on incident start/resolution timestamps. If you notice an inaccuracy or want to suggest an addition, contact us.