This is why too much centralization in ANY domain is a bad idea: programming, politics, finance, logistics, McDonald's ice cream machine repair technicians, you name it. Always, always, always have a back up plan.

Don't forget the 2018 fun: In less than a minute cloudflare changed a firewallrule worldwide to get rid of an exploit. Worked like a charm: 80% traffic drop. Which was caused by CPU-Overload in the Firewall. Whoopsie

We also had a similar problem last week here in Portugal. There was an outage at an Equinix datacenter in Lisbon, and multiple websites, including banks and public services, were down for a few hours.

17 days later, Rogers had BGP issues, and crashed 90% of Canadian stores, 9-1-1 services, and about 1/3 of Canadian's couldn't access the internet, or use their cellphones, home phones, or home TV services. Ah, what a marvellous world we live in.

Most of the new sites attracts to cloudflare because their free tier is super generous and they provide free SSL

I remember as a teen in the late 00s/early 2010s when I used to play a Neopets/Gaia mashup game named Subeta the site was constantly having issues with Cloudflare (the DNS service part iirc). Idk what it was about Subeta specifically that made Cloudflare act up or visa versa, but I remember that error page well.

funfact : Chess Olympiad was declared a draw after a widespread internet outage interrupted play in 2020 , making India and Russia joint winners. I remember seeing it live that day , it was intense.

Like or not, Cloudflare is the gateway for the Internet. Their services are just so good, it's difficult to justify not using them for web services. It'd be great if we could have more competitors and more options, but there's just no one on their level.

In Sweden we had a fun outage a few years back, I believe it was spring 2018. Some workers cut a fiber line while doing some construction work. Apparently this fiber line was the main trunk between the southern part of Sweden and the central part of Sweden. Absolutely nothing worked were I live. There was no cellphone or regular phone service since all pbx services are since a decade or a maybe more like a fee decades completely digital. Since there was no phone or internet there was no way to purchase anything without cash and we were unable to use our ATMs since they are connected to our banking system through the internet, or the phone network as a backup The reason for this is because the ATM always checks your account balance before you can withdraw anything (at least that's how I understand it). Talk about getting some idea of how it would be if we get a powerful sun flare😁 That for sure is something I'm NOT looking forward to. It took them about 10-12 hours to get the line repaired so it would work again, part of that was finding the cut because the workers who cut it hadn't noticed. The one thing Telia,(one of the biggest, if not the biggest ISP in Sweden and also the owner of the fiber) never cleared up is why the redundant connection didn't work at all; however, I've heard some people speculating that both cables actually got cut because the cut happened right outside the building with the ISPs equipment. However, looking at the options, they aren't that many. ~They cut both cables. ~The config where bad so the fail over simply didn't work. ~Faulty equipment that for some reason had not been swapped and because of that the fail over didn't work. Those are the options I can think of. Any other ideas folks? 😊

As DOTA player, I accept this LOL article as a challenge

I mean, I get that it's essentially a glorified load balancer, but this much centralization of the internet is a dangerous thing.

Cloudflare 🤝 AWS Taking half the internet down if they goof

2:09 - "A single point of failure for a huge portion of the internet". I'm sure this is a perfectly apt description of quite a few social websites out there... especially those that went dark recently, or smelling a tad musky.

On this same thread... an oops by Roger's took out Interac for several hours, causing Canadians to not be able to buy anything with Debit or Credit Cards across the country, as Interac may have had redundant data links... they were from the same ISP. It also took out internet for all their other customers too. Then few months later... another issue took out several provinces 911 services!

I love you trying to explain the weird things about my life and lively hood to people who don’t understand or have insight into how the internet actually works now

For many years, I always thought CDN meant "Canadian" because I live in, well, Canada. It took me some time to realize that it meant something completely different.

I was awake during this and it was CRAZY, every other website I tried was down with some ngix error popping up.

You could do one of these about the massive Rogers outage in 2022 as well.

The short steps for plugging in s single Ethernet cord hit home. Its exactly what ive been doing for some companies for the past several months, drive 4 hours, patch a single fiber cable, drive 4 hours home.

Can you make a video that explains what patch panels are?

Just in time for AWS us-east-1 going down.

I'm a little surprised that this video wasn't discussing the AWS outage last week.

Thanks for the video! I actually remember when Discord, LoL and others become down!

I love how you read their bio like an ad spot lmao

Nice job, Robert Oakes. I enjoyed the sneaky self-photo. 😉

I forgot that this wasn't an LTT sponsor spot. It really feels like one.

Quick criticism, a portion of your video was confusing as it was written. In the beginning of the video you mention an incident in 2022 then mention it again without the timing towards the end of the video. The problem is that your bridge to mentioning it in 2020 made it sound like it happened after the 2022 incident which was helped by not mentioning the time again the second time you mentioned the 2022 incident. We also saw this problem of too many eggs in one basket when AWS East-1 went down earlier this week, and it was having trouble for much longer. Cloudflare for only having less than a handful of major incidents I think is doing an excellent job. I personally use them and have very few problems (other than dynamic DNS updates. They only work for non-WWW subdomains for some reason.) The interface is incredibly easy to use and the documentation is pretty good.

Time to get that Mobo and Test the 4 Gen 5 M.2 slots Linus promised me at 2:31 . The picture says otherwise but I trust Linus! 🙏

Amazing, 1 thing goes down and everything goes poof. Thats what happens when everyone relies on that 1 thing... No matter how good that thing is, it will eventually fail. Even if its just 1 time every 10 years, disruptions will be hard to avoid. This video finally explains what Cloudflare is, i never knew and i kept on wondering. Techquckie doing its job

what LTT Video is that at 0:50 and what was the streaming service that made "an incomprehensibly bad naming decision"

Cloudflare also has some rather intrusive browser-checking systems for some websites. Not sure if they offer site hosting or just certain services for sites, but some sites that I used to frequent would have a delay when clicking to a new page as it did these intrusive "browser checks" so I just stopped using those sites.

And then there was the one time an employee was told to dismantle an old used rack, pulled one fibre out and killed the entire Cloudflare customer access😂

All I know is that Bitt shoot (happy, bot police?) requires their javascript to be enabled to watch videos there.

So no one is gonna talk about AWS outage 😅

It's gotten worse at least once - when Level 3 misconfigured one of their core routers and LITERALLY took down almost all of the Internet for a day and a half. When you could not even get through to GOOGLE (and Amazon), you know it was a BIG issue.

When this happened I called Comcast to report an outage cause it was stormy that day so i just figured that happened, found this out the next day when i had internet back. (i still use cloudflare

Did youtube upgrade their compression? The video quality is way better than it used too.

Cloudflare is a MITM attacker's wet dream.

Crazy that they are still in the same building in sf

Cloudflare argo tunnels are great for selfhosting

0:47 - Anyone know what this is referring to?

wow that is very dangerous to have these internet outages once in a few years. i mean people can actually end up reading a book

I just had a $million dollar idea! A litter box for dogs!!! Dogs would totally use it. They would understand the concept so quickly!

Textbook example of the downside of the TCP/IP networking standard, it's actually "decentralized" (notice the quotes.)

Cloudflare is just so good.. I mean I dont have to pay for my DNS Records or my DDOS protection.. as a developer with no money its great hosting my domain for free

Sure would be cool if they made a V4 Predator 😂. Although you can get lightly used 2.0L 'Japan spec' Subaru motors for $600.

Using dna benchmark they are number one with my dsl co but cable there local dns was better but I have even better with pfsense local dns except for when it goes searching it’s 500ms slower then cloud flare

Haha, I have reflex to tap for fast forward when I hear “League of Legends” 😂

Which streaming service recently made an 'incomprehensibly bad naming decision"? I mean, I know Floatplane is a weird name, but it's not that bad...

That's not a Monday detail .Michael!!!

How about a Tech Quickie on GitHub, etc....? 🤔

4:26 - if that list is sorted by idn , smells like an off-by-one error, lol!!

I actually watched your sponsor because: Starfield. 😂

Too much centralization of the internet. Almost every domain is heading in this direction. 😢

That Linus segway makes me want to change DNS, Browser and ad blocker all at once

No Riley, “DNS” is short for *”Domain Name Server”*

Isn't it "Domain Network Service"?

Dan can be helpful but also scary

oh this was a video? look like a infomercial for cloudflare.

Because of that, I'm using my own nameservers and cdn

I hate seeing ads in techquickie.

Why didnt I hear anything about an internet outage last year?

Oh…so it was accidental few I thought it was something else like, oh, I don’t know ON PURPOSE!

Aaah, so almost everything runs with Cloudfare. Great... I'll remember this tip when needed.

"There goes comcast" *Ding*

I really hate that TrashFlare

Did Cloudflare's outage affect any backbone or only sites using their servers?

Putting my IPv6 server behind cloudflare make it accessible from IPv4

Wasn't there a similar issue somewhere in the past were many Google services were down due to an accidental "/" that was added in the code?

So cloud fare was why my discord made me do phone verification for no reason

4:38 had me dying 😂😂😂

DNS = Domain Name Server, Domain Name Service or Domain Name System? I have literally heard and seen all of them in all kinds of different networking stuff. XD

Was the incomprehensible naming decision Floatplane?

Now its Cloudstrike that went down and took half the e World with it 😂.

This is pretty good

You guys could've explained the issue a little better than a single line of code. I get that it's supposed to be a quick video, but you should explain that the internet runs on routing protocol EBGP and that it's not really a dynamic routing protocol therefore routes must be added manually, in the correct order, and redistributed routes may need to be filtered. Yes DNS can wreck havoc as well. People don't understand that the internet, or the so called "cloud" is NOT PFM, and that there is no such freaking thing as a cloud. It's someone else's servers in someone else's data center plain and simple.

Does DNS still need to be explained in every video?

oh, i was told it was AWS servers that went down that caused all the issues that month.

Only thing possibly faster than Techquickie is the comments

For a moment I thought this was a cloudflare ad.

My brother's company uses cloudfare for its websites and he was not happy when it went down. 😂

Now look into akamai and fastly

They mess up a pointer in the code a pointer shuts the the internet

Well that company is now officially my Enemy.

I would love to see techquickie come to spotify, after all they can be easily enjoyed without video. This would make listening offline easier for many people.

Cloudflare also has a lot of other services you can use for FREE

I agree , Internet must be down, as soon as posibble. People be manage reality with real close physical activity, for better future,

Sure !

Do a Techquickie on LoL, and how toxic that is haha. I dare ya, cus it's probably impossible to do it fast. Talk about a deep rabbit hole.

We use something similar, and to be honest trying to do this stuff on your own over almost 30 countries will cause more downtime than any of these big CDNs. Mark my senile gibbering sleep-deprived words.

even i use cloudflare for my personal website

I thought DNS is domain name Service

Thank you Cloudflare Warp for letting me play online games on College Wifi

Why didn't you guys talk about the Roger's outage back in 2022? This was bigger then Cloudflare's network outage...don't get me wrong it's massive, but an entire ISP going down for somewhat near 36-hours was even bigger.

How can one be so amazingly likeable as Riley is? I just don't get it. He is such a great guy! Love you Riley ❤ (mostly platonic 😜)

wow, a video by Riley without those cornyass jokes? it's now becoming better. good job 👍 keep it up.

"One day, a planet, called Earth, went down, and took all the internet with it. So what even is "Earth?"" This is what these videos sound like, half the time. But only for us IT non-professionals.

It's also fun when any major company has a service outage (looking at you, Microsoft). Glad my job doesn't rely on any single thing these days, but it was significant some years ago. Also, enjoying the "Digitall Power Design" - smoothly transitioned off that slide in a hurry. Not sure if it was actually a typo, though.

You have to give them credit for screwing up in 2022 and doing it again in 2020.

Ah, I remember that happening^^ I think it was a DNS problem Edit: nope, seems like I've mistaken it for that facebook outage where they were in a local network not accessible from outside or something like that