Why the heck is everything stripped to them. I can’t believe the grandfather’s of the internet thought about a www which goes completely dark if some players like cloudflare, aws, azure etc went down. Put down these network oligarchs!
I run a nonprofit and as soon as I got our new website up, a board member insisted the domain route through Cloudflare. I choose my battles with this guy, so I went ahead and did it. Thus, fixing a nonexistent problem means our website is now subject to Cloudflare downtime.
Might be a good moment to ask the Board if they’d like to go another route
Easier and cheaper to outsource your site’s security to someone else. God forbid you have to learn the cyber. Simple as.
Oh, let me check down detector to see what’s impacted.
Well, shit…
Man what’s not impacted?
I host my own shit and I couldn’t access it from my own houseUse a tailscale tunnel instead, or at least as backup.
I have wireguard, but I don’t expose most services locally, so it wouldn’t save me in this case.
Thanks though
as a general noob to self-hosting, I’m curious about this!
do you run your stuff through a cloudflare tunnel, or have any authentication service that lives outside of your network? something has got to be causing your connections to go out and then back for a cloudflare outage to impact on your own LAN, right?
Maybe cloud flare dns 1.1.1.1?
DNS? DNS.
Pork rinds? Pork rinds.
Relaying on one big entity is always a bad idea, self host powers
TFW Cloudflare is down but I can’t tell Lemmy about it because… Cloudflare is down (and Lemmy.world uses Cloudflare).
Oh, I thought my eduroam was just being shit today, that’s good to know
God, screw this stupid roadblock of a nuisance. I have to go through this piece of shit everytime I need to log-in to my bank account’s app. Sometimes it’d even fail and I won’t go through. Even if I did go through, I have to double-verify anyways! I’m so over this stupid era of verifying yourself multiple times.
Funny when downdetector is down because they use cloudflare
State actors: Let’s take down . . Ohhh . . Howabout cloudflare today.
Paranoid people: State actors! Technical people, engineers, programmers, normal people: shit happens, people built this.
Just because you’re paranoid, don’t mean they’re not after you
I’m guessing it was actually something internal. If you look at their status page you’ll notice the outage occurred smack in between some sort of maintenance work they seem to be rolling out to most/all of their edge locations. As soon as they resolved the outage they continued with the regional maintenance updates.
Of course known maintenance windows are the perfect time for the ol’ cyberattackeroo.
let me update my notes… one second
I’m guessing it was actually something internal.
My sources tell me it was a Russian egg sac that hatched early and shorted out some power rails. It’s plausible, it certainly can’t be the house of cards nature of modern systems.
Good thing that the dezentralized internet has a centralized arbiter that…if down, effectivly, gives the entire internet a single point of failure. As cloudflare seems to have a use, especially in modern times were everyone and their mother got an ai crawler sucking small servers dry, I wonder if there is anything left to have interconnected networks without some big shot buffering against accidently ddos
“Further details will be made available when we’ve worked out how the AI screwed us this time”
Is AI the new DNS?
switches over to stored media
…again
CloudFlare went from the muscle of the internet to a liability.
I wonder if it’s even worth it now.
Might be better to try and wing it on your own if you want reliability.
Gov and corpos will keep using it since it absolves them of responsibility.
Today, we remind you that AWS isn’t the only single point of failure on the Internet.
Multiple customers.
Yeah, 8 billion is multiple customers
Potentially
The Chinese United Front are testing a series of backbone organizations and assets for readiness. The final attack will coincide with major world events and obscure readiness and communications.
This will be done via disconnecting several relates levels of internet functionality so that even their repair measure will still seem unresolved. They will reconfigure and retry, unaware that their problem is solved, it’s the other levels/locations of the internet’s function that are still broken (so can’t be tested against).
This will provide extended cover that could last days or weeks. So far testing of this plan has been going well.
