A global outage on Cloudflare, a network which offers web application security and performance, has had an outage today, which began at around 11:30 GMT. This has affected many websites, including our own, with users seeing Cloudflare error pages rather than the actual webpage itself.
Other services, including X.com (formerly twitter) and chatgpt.com have been affected. Cloudflare has been working on resolving the issue, which looks to be resolved for some services as of 15:00 GMT, but some issues still remain. We saw our website restored fully at around 12:50 GMT.
We’ve had a look through our broadband provider list and noticed that a number of broadband provider websites or broadband related services were affected by the outage due to using Cloudflare on their own homepages. This doesn’t mean their Internet service was affected. This includes:
- Airband
- Community Fibre
- Country Connect
- County Broadband
- Exascale
- Fusion Fibre
- Grain Connect
- Hampshire Community Broadband
- Hey! Broadband
- InternetTY
- Jurassic Fibre
- LightSpeed
- Lit Fibre
- Pulse8
- Quickline
- Trooli
- Wildanet
- YouFibre
- TOTSCo
We have also seen reports from broadband providers reporting increased call volumes from customers about the problems, unsure if the issue was related to their broadband service. Alex Bloors, GM of AAISP, a broadband provider focussed on more technical customers, confirmed even they had seen calls from customers. Community Fibre had added a banner on the top of their website, presumably having received many calls on the issue:

Users visiting affected websites will have seen a familiar 500 Internal Server Error:

If you are having an issue reaching your broadband provider through their website we’d suggest giving them a call. (A Google search should bring up a phone number if you can’t reach the website itself).
I started seeing problems well before 11:30am
I agree, I first noticed when a colleague said a website was down and I asked them to try mine, seconds later it was Cloudflare proxy errors. Around 1 hour later things really kicked off, downdetector down (how ironic) and more and more platforms went offline.