Previous incidents

API recovered.

This was largely a continuation of https://status.craftcms.com/incident/620811?mp=true.

A combination of aggressive bot traffic (AI and normal) managed to make its way through our firewall and rate limiting logic in the EU region, affecting some shared infrastructure resources.

We’re working with Cloudflare to find the root cause of the issue and discover why it’s not happening in other regions, and are doing this while trying not to affect legitimate bot traffic.

In the meantime, we’ve im...

Around 6:15 AM PT, we started to receive reports of performance issues and timeouts for some sites in the EU region.

The issue was resolved around 6:30 AM PT.

The underlying issue was a flood of DDOS traffic that made it through our firewall and affected some infrastructure in that region.

We have increased the resources available in that region to help mitigate situations like this in the future.

We have adjusted our global firewall rules to help further mitigate future issues like this....

Spoke & Chain and Europa Museum recovered.

DNS & Gateway recovered.

Spoke & Chain recovered.

The issue appears to be with routing on the Cloudflare side and only affected some Craft Cloud sites that use DNS A records pointing to our Cloudflare IP addresses instead of CNAME records pointing to edge.craft.cloud.

Will post more info as we continue investigating.

v1.62.0 of the Cloud API introduced a regression where specific node and craft-cloud.yaml configurations would not get their artifacts published to the CDN during a build.

This is resolved in v1.62.3, which has been deployed to all regions.

In order to mitigate issues like this going forward, we are expanding our integration test coverage to cover...

Cloudflare has posted a post-mortem of the incident here: https://blog.cloudflare.com/cloudflare-service-outage-june-12-2025/

Spoke & Chain went down.

Europa Museum recovered.

The issues with non-production subdomains improperly routing has been resolved.

Some customers in the European region are experiencing intermittent database connection issues.

The issues have been resolved.

Further investigation will follow.

We have identified the cause and have deployed a fix that resolve this issue and prevent it from happening in the future.

Diagnosis

In the event of an unstable network connection between Cloudflare's servers and Craft Cloud's origin servers, combined with a threshold of traffic – there was the possibility that responses could be cached in an incomplete (or empty) state.

The unstable connection is not a issue in itself – network fluctuations are inevitable and need to be accounted for.

...

Europa Museum recovered.