Another vote to confirm that AWS appears to be working fine. At least EC2 and S3 which is basically all we use (highly recommend this simple setup using base primitives, as some of their higher level services have significantly lower reliability)
Downdetector is end-user-reported, and the numbers look reasonable, so I'd say it's doing fine. I think more likely the cloudflare outage is causing failures that end users are reporting as outages in the service cloudflare is supposed to be proxying.
Update - We are seeing a number of services suffer intermittent failures. We are continuing to investigate this and we will update this list as we assess the impact on a per-service level.
Impacted services:
- Access
- WARP
- Durable Objects (SQLite backed Durable Objects only)
- Workers KV
- Realtime
- Workers AI
- Stream
- Parts of the Cloudflare dashboard
- AI Gateway
- AutoRAG
- Jun 12, 2025 - 19:02 UTC
Update from Cloudflare: Identified - We are starting to see services recover. We still expect to see intermittent errors across the impacted services as systems handle retried and caches are filled.
Jun 12, 2025 - 19:12 UTC
I’ve been telling people this is a terrible idea since many people on this site were literal babies.
I’ve also been telling people centralized OIDC identity owned by a couple oligopoly vendors that have no incentive to care about normal users is a terrible idea for the same length of time.
Unfortunately convenience is the most powerful force in computing. People will trade security, privacy, autonomy, resilience, ownership of their own data and even ownership of IP, and vendor neutrality (avoiding lock in) for something convenient that “just works.” 90%+ of people will make this trade every single time.
It’s a big reason that FOSS never makes it outside techie circles or the back office / cloud. The cost and difficult of making a product easy to use for regular people increases exponentially as you go up the “easy-ness” gradient. Only businesses with real money can afford to make software usable.
Often making software usable is many times harder and more labor intensive than making the base software work. This is because computers are actually very hard to use and confusing. If you don’t agree it means you are an expert and don’t see the difficulty any more, in the same way that e.g. an athlete might not understand what’s so hard about difficult physical tasks.
It's interesting that internet infrastructure is so interconnected that failures of what _should_ be separate systems seem to coincide.
Do we think it's possible the GCP outage caused the Cloudflare outage, or vice versa?
For DownDetector, I know it's not exactly foolproof. I imagine a lot of sites use something like CloudFlare in front of AWS infra, which could easily be automatically flagged as an outage for both in their system.
Maybe it was my lucky day, I was not affected at all. Every platform I used was totally fine, even deepseek, which uses cloudflare for verification, worked fine.
49 comments
[ 4.5 ms ] story [ 102 ms ] threadhttps://news.ycombinator.com/item?id=38987701
https://news.ycombinator.com/item?id=38939668
https://health.aws.amazon.com/health/status
https://azure.status.microsoft/en-us/status
Update - We are seeing a number of services suffer intermittent failures. We are continuing to investigate this and we will update this list as we assess the impact on a per-service level.
Impacted services:
[1] https://www.cloudflarestatus.com/I’ve also been telling people centralized OIDC identity owned by a couple oligopoly vendors that have no incentive to care about normal users is a terrible idea for the same length of time.
Unfortunately convenience is the most powerful force in computing. People will trade security, privacy, autonomy, resilience, ownership of their own data and even ownership of IP, and vendor neutrality (avoiding lock in) for something convenient that “just works.” 90%+ of people will make this trade every single time.
It’s a big reason that FOSS never makes it outside techie circles or the back office / cloud. The cost and difficult of making a product easy to use for regular people increases exponentially as you go up the “easy-ness” gradient. Only businesses with real money can afford to make software usable.
Often making software usable is many times harder and more labor intensive than making the base software work. This is because computers are actually very hard to use and confusing. If you don’t agree it means you are an expert and don’t see the difficulty any more, in the same way that e.g. an athlete might not understand what’s so hard about difficult physical tasks.
Do we think it's possible the GCP outage caused the Cloudflare outage, or vice versa?
For DownDetector, I know it's not exactly foolproof. I imagine a lot of sites use something like CloudFlare in front of AWS infra, which could easily be automatically flagged as an outage for both in their system.
Although my Racknerd VPS is doing fine.