📈 📵 Major internet outages involving X / Twitter, Cloudflare, Target, Open AI, Amazon Web Services, and Microsoft. 📵📈
(www.thousandeyes.com)
NET FAGGOTRY
Comments (39)
sorted by:
We had that post yesterday about the Simpsons predicting wide scale outages and today we have this, interesting.🤔
Cloudfare has an outage like this weekly
Not really; "the machine" (Claude in this case) crawled the web for 10 minutes and then concluded: While Cloudflare does experience occasional issues, truly global outages that take down major websites for hours are not a monthly occurrence.
Not that I have noticed. In Canada, darn near half my content was unavailable.
10 minutes of darnkess
Felt like 10 days to me but that's just me
Way more than 10 for us - several hours
take your upvote and gtfo 😆
Cloudflare System Status - https://www.cloudflarestatus.com/
Also prevented access to greatawakening.win and scored.co...
It prevents access to any cloudflare site which is easily 1/3rd of all sites
I was able to get on by changing my VPN a few times till I found one that worked. I primarily use ProtonVPN but when I have issues with that one, I switch to my PIA account. I also have a dedicated IP via PIA. Use that one for banking and such.
Jesus these posts irritate me to hell and back. How is this worth a sticky on this site??? Outages happen. Humans are stupid and humans are the ones running these systems.
Devops engineer here. Outages are not news. Even the massive us-east-1 outage a few weeks ago, was just some nitwit engineer who flipped a value incorrectly and brought a whole region down.
Theres no instance where a large outage will be required to release or gather intel. In fact, if anything, a large outage would be the opposite - it would be to prevent something getting released, but even that would be very temporary, and likely accidental.
The reason it always seems so widespread is because of the interdependence of these systems and developer incompetence. If a site filters your traffic thru cloudflare and cloud flare goes down, you lose the entire site (ALL sites that use cloudflare). My infrastructure has been fine thru every aws outage and cloud flare outage because im aware of these issues and didnt build dependencies into my systems where someone else's fuck up could bring my shit down.
Tldr: outages arent news. It being stickied here is a waste of time, space, and attention. And yeah im aware of a delta today. Doesnt change anything considering cloudflare has some sort of massive outage WEEKLY
I agree that most outages have a few hundred people reporting no service. No big deal.
This one was much bigger than most. Downdetector was stuffed full of scores of sites (just like you said, due to cloudflare) that were down with upwards of 15,000 individual reports (That's JUST people that know how to report outages).
My wife couldn't get into her school system, my coworkers in Florida were dark, this site went down MOMENTS after I made this post. X and Truth Social were also out.
This was an anomaly.
BUT a significant anomaly one day after a prediction of one!
It actually wasn't significant at all, as my entire comment is implying. Wasn't even all down for an hour lol
If an outage is serious serious, like global new world order restructuring and whatnot, we arent gonna be here discussing it on this forum. A real outage with real, world changing consequences will be painfully obvious when it happens. People will die, systems will be offline for Days/Weeks, planes will be grounded, nukes will be armed, travel halted, likely martial law instituted. Trust me - if an outage is sticky worthy, it wont need a sticky here for us to know about it. Thats all my point is
No disagreement, I said “significant” (and more with regard to the timing), not “serious”.
It did last about 2-4 hours here, but that’s not even “call your doctor on Viagra” range. Iirc there’s been some outages that knocked a lot of systems out for close to a day before, and even with that, most things keep chugging along.
I was 2 hours from home in a large city and Apple Maps went down. Unfamiliar city with no navigation. I was relieved when it came back up.
Now that's scary!
That happened to me a couple years ago while driving through Harlem with the family. Never happened before and definitely had me spooked.
Umm point of order - If you're running a third of the globe's Web traffic through your servers and your company's regular service dies for hours for all of your customers, it's fucking newsworthy. It certainly was for the amazon outage last month, and cloud flare is yet another one of those major providers who are now close to core infrastructure as a piece of software for most of the Web.
I've been in tech and worked in devops for a while now. If something goes down for your company, do you keep quiet about it to your customers? Not in my line of work. You flag the issue and someone (usually the product owner/Manager aka a guy like me) presents an scr to the higher ups, or puts out a notification on some SocMed site that works over what's occurring and how long it will be until it's fixed. This is why most tech firms have a status page if they have any live services.
The fact CF borked its tech globally for its entire customer base for hours today is not a nothing burger to me. It was a highlight on the frailty of the net - give aws and CF a hard enough kick in the nads, and the whole Internet will collapse into a broken, squealing mess.
Also CF doesn't go down for all customers for hours a week. If it did, no one would use them. Hell, their cto probably would have been fired if he let that level of derp go out in his codebase regularly. I sure as hell would be kicking him to the curb if that were true and I was a CEO there. The buck stops with them, and last I checked Dane Knecht was still in charge of the situation. If he was presiding over a weekly omgwtf moment like this morning was in London for a few hours, he'd have been let go so fast that Elsa wouldn't finish the first chorus before he hit the pavement.
There is no way in the world that you can prove this point. Looking back at the history of Q on the chans, one might even call it naive.
Well said. It’s funny when these things happen everyone is like omgz it’s happening !!! More to your point, events like this would put more eyes on systems making it more difficult to get things out. There are some memes/likely theories that most recent outages have to do with poor automation/people relying on AI to deploy changes/updates. …. ChatGPT push these changes to prod. Make them scalable. Oh and make sure they don’t crash the environment. wipes brow.
Aaaaaand it's back!
Qioncidence?
This morning, Mike Johnson was discussing the upcoming Epstein house vote to safely release the files.
Cloudfaeries....
All morning...no .wins
u/#catdance
BE CAREFUL....MCROSOFT IS GETTING HAMMERED....that is a good point to inject MAJOR damage to a company. If hackers were able to offloaded corrupt packages into a distribution payload and deliver that to every updates windows package....that would stop more than you know
But what do I know
Yup. CloudFlare was inaccessible for about three hours this morning. I was wondering if it would come back up at all, but here I am.
Damn it. I need the AIs for a protective order against my ex wife. Fucking outages.
you can do those online?
You can use an AI to draft the motion/application. In Texas you can efile them online. But you need to really take a PO to the court. The judge will want to hear from you that day in order to grant an Ex Parte order. Meaning with out the other person being there. Then they will set a hearing 14 days out.
I suggest using Claude.ai for motions, it under stands stuff better than GPT. But to do have GPT double check Claude.
Hope this helps. I'm using the AIs to do all my pro se filing and saving a load on lawyers. But you have to do research and ask them loads of questions.
If you go to the court corridentor they will be tight killed about how to do things, be sure to tell them you're not asking for legal advice just how procedure works at the court.
It's not as easy as you might think and the AIs do fuck up some times.
Good luck!
Yeah, I figured this was the case after the site went down for hours
Funny, this site was down this morning.
I'm in Europe... our morning was normal, Cloudflare had a service degradation starting around noon UK time / 12:00 UTC.
Definitely a lesson in not putting all your eggs in one basket, even though Cloudflare has historically been reliable, and generally putting a service behind their CDN/Proxy/Firewall solves a lot more problems than it creates. (DDOS protection, HTTPS termination, caching, and content distribution to speed things up for far away users)
"Definitely a lesson in not putting all your eggs in one basket"
Or engineering single points of failure for the entire internet...
Bank of America bill pay was down, also - for me at least.
My credit union was down for a few hours this morning.
Had so much trouble getting anything done online.