Overview

• Google Cloud engineers traced the outage to an IAM malfunction affecting over a dozen GCP products including Cloud Data Fusion, Memorystore and Cloud Storage and have deployed targeted fixes.
• Cloudflare identified cascading failures across services such as Access authentication, Workers KV and Zero Trust WARP and reported that traffic began normalizing as caches refill.
• Users may still encounter sporadic failures while systems process queued requests and refill caches over coming hours.
• Downstream disruptions affected platforms including Spotify, YouTube, Discord, Twitch, GitHub and Mailchimp, generating tens of thousands of reports on Downdetector.
• The outages spanned multiple regions from North America to Europe, Asia and Australia, underscoring the vulnerability of centralized cloud infrastructures.