Skip to main content

in reply to along_the_road

TL;DR: React broke the internet.

Well, that, but also Cloudflare went down because they were trying to fix React's shit.

in reply to Midnitte

That's an HTTP status code, I know what those are :P
This entry was edited (16 hours ago)
in reply to Lena

When laypeople see "404" I think they generally think of something other than "status" code, but I suppose you are correct.
in reply to Midnitte

I think it also doesn't help that only 4XX (client error) and 5XX (server error) are defined as error status codes, and 4XX errors don't even necessarily indicate that anything happened that shouldn't happen (need to reauth, need to wait a bit, post no longer exists, etc).

Trying to think of what 6XX would stand for, and we already have "Service Unavailable" and "Bad Gateway"/"Gateway Timeout", so I guess 6XX would be "incompetence errors". 600 is "Bad Implementation", 601 is "Service Hosted On Azure", 602 is "Inference Failure" (for AI stuff), and I guess 666 is "Cloudflare Outage".

in reply to along_the_road

Stop using it already. The internet was not meant to be centralised.
This entry was edited (1 day ago)
in reply to Lime Buzz (fae/she)

On the one hand, I 110% agree with you

On the other hand, it's so damn convenient. They cache your shit and they protect you from DDoS attacks, and they do it for free*

*Until you're big enough to warrant extortion from them.

in reply to boonhet

I am pretty sure that 99% of sites would have less downtime due to DDoS attacks than from such outages. I have so many issues with Cloudflare that I don't even know where to begin with, from over-caching causing issues up to decrypting all traffic, who the hell thinks this is really a good idea?
in reply to along_the_road

I like that the headline needs to include the date so people know this is not an article from a few weeks ago.
in reply to along_the_road

Is there a reason these outages seem to have increased recently?
in reply to Megaman_EXE

Without looking into this specific outage, I'd suggest things like deferred maintenance and "cost optimizing" technical staffing are often contributing factors. (At least in my experience)
This entry was edited (1 day ago)
in reply to Megaman_EXE

Lack of NSA funding to run their man in the middle platform that everyone likes.
in reply to Megaman_EXE

This entry was edited (1 day ago)
in reply to t3rmit3

This is the actual answer with respect to Cloudflare. Their config system was fucked in November. It's still fucked in December. React's massive CVE just forced them to use it again.

More generally, the issue is a matter of companies forcefully accelerating feature development at the cost of stability, likely due to AI. This is how the company I'm at is like anyway.

in reply to Megaman_EXE

Is there a reason these outages seem to have increased recently?


We've had three years of unnecessary tech layoffs.

Nobody knows how the fuck anything in their technology stack works anymore.

Everyone is just spinning the giant wheel over and over and hoping it doesn't land on bankrupt.

Sell your technology stocks, kids.

in reply to along_the_road

There is now a blog post from cloudflare on the outage: blog.cloudflare.com/5-december…
in reply to along_the_road

It seems my moving away from major tech companies is working well, I didn't even notice there was an outage (again).