There seems to be similarity to the AWS attack a few weeks ago.

If you're struggling to load posts on X, create work on Canva, or play a few games of League of Legends or Valorant, then this is likely why. Generic messages like "Please unblock challenges.cloudflare.com to proceed" appear on several websites, including OpenAI, which blocks access to ChatGPT.


Started at 0000 Zulu (6am CST for me).

Ventusky, some news sites, & CPA Academy online continuing education courses are down.

Nearly unable to work today at present due to the outage.

How can ONE "site" bring down the internet?

Isn't anyone minding the store?

Guess it is time to take Sandy The Little White Dog for her morning walk, even though it is freezing outside.

Thanks for explaining. The other main forum I participate in is down as well but good ol’ SF keeps plugging along!

2 of my usual forums, 2nd to here are also down.
Our HSE site at the office is down as well.

Yep,my local forum is down, same with my streaming service (CinemaHD)

My second most visited forum (talk bass dot com) is down! All hail Para!

I think it's working now.

quote:

Originally posted by Ripley:
I think it's working now.

Just checked the 2 down forums & they're both back up.

Looks to be still having some fits, but improving. One of the sites is giving intermittent drops.

ohthehumanity

My brokerage's website uses Cloudfare so they sent a note that are data is safe, but we can't view it until Cloudfare comes back.

I'm not a day trader so no big deal

quote:

Originally posted by 4MUL8R:
How can ONE "site" bring down the internet?

There's truth in the old saying, "If builders built buildings the way programmers write code, the first woodpecker to come along would destroy civilization."

quote:

Originally posted by joel9507:

quote:

Originally posted by 4MUL8R:
How can ONE "site" bring down the internet?

There's truth in the old saying, "If builders built buildings the way programmers write code, the first woodpecker to come along would destroy civilization."

I suspect that it was more a "too many eggs in too few baskets issue. Cloudflare isn't a "site," it is a service. They host many websites and perform this thing called "anycasting" so that these "sites" located around the world can all serve the same content to queriers located topologically "close" to them. E.g. a web surfer in SoCal might get their content from a server farm in Nevada, while a surfer in NYC looking at the same content would be getting it from a server farm in NoVA. This cuts down significantly on long-haul Internet traffic, and generally improves site performance from the POV of the querier. This anycasting is done on a router level, not in the web servers themselves or any other endpoint.

So, until there is actual analysis of the outages, I will assume that some group gained control of some of Cloudflare's router infrastructure, or managed to inject control messages that broke their anycasting.

Here is Cloufflare's initial report on the root cause of the problem. They attribute it to a "misconfiguration," but whether this was caused by a malevlent intruder or ham-handed network administrator is left unsaid. It appears that the outage may have cost Cloudflare tens of billions of dollars. Oops, wrong event, the URL points to a report on the 7/17/2020 outage, not today's which appears to be forthcoming.