[madjam002]

Looking forward to the post mortem on this one. We weren't affected (just using the CDN), and people are saying they weren't affected who are using Cloudflare Workers (a previous culprit which we've since moved off), so I wonder what service / API was actually affected that brought down multiple websites with a 500 but not all of them.

Wise was just down which is a pretty big one.

Also odd how some websites were down this time that previously weren't down with the global outage in November

[archon810]

Our locations excluded from Cloudflare WAF were up, but the rest was down. I think WAF took a dump.

[reassess_blind]

Yeah it's strange. My sites that are are proxied through Cloudflare remained up, but Supabase was taken offline so some backends were down. Either a regional PoP style issue, or a specific API or service had to be used to be affected.

[da_grift_shift]

The excuse:

>A change made to how Cloudflare's Web Application Firewall parses requests caused Cloudflare's network to be unavailable for several minutes this morning.

>The change was deployed by our team to help mitigate the industry-wide vulnerability disclosed this week in React Server Components.

>We will share more information as we have it today.

https://www.cloudflarestatus.com/incidents/lfrm31y6sw9q

[madjam002]

It's quite an unfortunate coincidence that React has indirectly been the reason for two recent issues at Cloudflare haha

[GoblinSlayer]

Who sent an xml request?

[gritzko]

The entire Cloud/SaaS story had a lot of happy-path cost optimization. The particular glitch that triggered the domino effect may be irrelevant relative to the fact that the effect reproduces.

[thinkindie]

we were not affected too and we realised it was Cloudflare because Linear was down and they were mentioning an upstream service. Also Ecosia was affected, and I then realised they might be relying on Cloudflare too.

[themly]

CDN was definitely down also. We were widely impacted by it with 500's.

[gowthamgts12]

CDN was also affected for some customers. we were down with 500.

[m_mueller]

Maven Repository was down for me for a while, now it recovered.

[cryptonym]

> Looking forward to the post mortem

This is becoming a meme.

[meandmycode]

This has to be setting off some alarm bells internally, a well written postmortem on an occasional issue, great, but when your postmortem talks about learnings and improvements yet major outages keep happening, it becomes meaningless..

[kryptn]

was interesting, some of our stuff failed, but some other stuff that used cloudflare indirectly didn't.