A strange sudden outage has hit several major online services today. Google Cloud and Cloudflare directly reported global issues at around the same time(17:00 UTC), with some services already back online(such as cloudflare’s turnstile and WARP) but many still recovering(Google’s API Gateway, Agent Assist, AlloyDB for PostgreSQL, etc.
While Microsoft and Amazon have not talked about any issues with their Azure and AWS services, it is expected that those were also affected as both Microsoft Outlook and Twitch, which are hosted on those, were also affected.
Other services like Spotify, OpenAI, Discord and video platform Vimeo were also affected, though likely due to being hosted in the above mentioned cloud services.
There isn’t much information at the moment as to what caused such global outages, but engineers from those companies state in their status pages that the “root causes have been indentified and mitigations put in place”.
Some users speculate that the incident might be related to storms going through the U.S. however that theory is unlikely as global services would not be affected by a single country having issues.
Update 21:37 GMT-3: Cloudflare just made a blog post explaining that on their side, the cause of the outage was “due to a failure in the underlying storage infrastructure used by our Workers KV service, which is a critical dependency for many Cloudflare products and relied upon for configuration, authentication and asset delivery across the affected services”
The company denies this event was caused by an attack, saying that its root cause was an issue with a 3rd party used by them for said storage infrastructure, however it remains a mistery how a single 3rd party provider could affect multiple major cloud providers.
Update 13/6/2025, 19:00 GMT-3: Google claimed “the issue occurred due to an invalid automated quota update to our API management system which was distributed globally, causing external API requests to be rejected. To recover we bypassed the offending quota check, which allowed recovery in most regions within 2 hours. However, the quota policy database in us-central1 became overloaded, resulting in much longer recovery in that region.”
Amazon(AWS/Twitch) and Microsoft(Azure/Outlook) have yet to make any statements.
This article will be updated as more information comes out.
Sources: Google Cloud Status, Cloudflare Status, Twitch Status, Down Detector, Cloudflare Blog
