When the Internet Went Down, Ever Studios Clients Stayed Up - Ever Studios Blog

0:04

Yesterday, the internet went dark, apps

0:06

froze, websites crashed, and millions of

0:09

people lost access to their favorite

0:10

internet services. What caused the

0:13

outage? A failure at Amazon,

0:15

specifically Amazon Web Services or AWS.

0:20

AWS allows internet companies to rent

0:22

server space and run their services.

0:24

Yesterday, there was a glitch at AWS. As

0:27

a result, more than 1,000 companies went

0:29

offline. The disruption lasted a few

0:32

hours and it triggered some key

0:33

questions. Apps and websites are central

0:36

to how we experience the internet. When

0:39

they stop, daily life stops with them.

0:43

So, should so much power sit with one

0:46

company or a handful of companies? Some

0:48

American lawmakers say that this should

0:50

change, that the internet should not be

0:52

so dependent on a few tech giants. Our

0:56

next report tells you about the debate

0:59

that is playing out and how AWS broke

1:02

the internet.

1:03

Yesterday, the internet stopped working.

1:07

Apps froze, websites crashed, and

1:10

millions of people couldn't get online

1:13

for a few hours. It felt like a digital

1:16

blackout.

1:18

The reason, a failure at Amazon Web

1:20

Services or AWS.

1:24

More than a thousand companies were hit.

1:26

Snapchat, Reddit, WhatsApp, even online

1:30

tax portals and banking apps. All of it

1:32

because one company had a bad day. AWS

1:37

is the world's largest cloud company.

1:40

Think of it as a giant warehouse for the

1:42

internet. It stores the data, pictures,

1:46

videos, and software that power our

1:48

favorite apps. Instead of every company

1:52

buying its own servers, they rent space

1:54

on AWS.

1:56

So when AWS breaks, the internet breaks

1:59

with it. Yesterday morning, that's

2:01

exactly what happened. Amazon's

2:04

engineers noticed increased error rates

2:07

in one of their main data centers. It

2:09

was based in North Virginia, a key hub

2:12

for AWS.

2:14

Soon, users across the world began to

2:17

lose access. The cause, a DNS failure.

2:21

DNS stands for domain name system. Think

2:25

of it as your phone's contact list. You

2:28

type and your phone dials the right

2:30

number. The DNS does the same thing, but

2:33

for the internet. It takes a website's

2:36

name, like google.com, and finds its

2:39

location online. But yesterday, that

2:42

system stopped working. It couldn't

2:45

match names with addresses. So when you

2:47

opened an app, your device didn't know

2:49

where to go.

2:51

The app was still there, but your phone

2:53

couldn't find it.

2:55

Experts say DNS failures are common.

2:58

Sometimes they happen during

3:00

maintenance. Sometimes it's a small

3:03

configuration mistake.

3:05

Sometimes it's a chain reaction inside a

3:08

data center.

3:10

In this case, AWS blamed an internal

3:13

error, a glitch in one of its database

3:15

systems. That single point of failure

3:19

spread across the network and within

3:21

minutes, a large part of the web went

3:23

dark.

3:25

This outage exposed how fragile the

3:28

internet has become. Most of the world's

3:30

apps run on the services of just a few

3:33

companies. They host everything from

3:35

emails to payment systems. is cheaper

3:38

and faster for businesses, but it also

3:41

means a single failure can cause ripples

3:44

worldwide.

3:46

There are smaller players like IBM,

3:49

Alibaba, and Stack It in Europe, but

3:52

none match AWS in scale. Replacing

3:56

Amazon's network would be like trying to

3:58

replicate the world's electricity grid

4:00

overnight. That's why the outage has

4:03

reignited a political debate.

4:07

US Senator Elizabeth Warren posted a

4:09

sharp message after the blackout. She

4:12

wrote, "If a company can break the

4:14

entire internet, they are too big.

4:17

Period. It's time to break up big tech."

4:20

Her post followed a string of antitrust

4:23

cases in the United States.

4:26

Earlier this year, a federal judge

4:28

refused to make Google sell its Android

4:31

or Chrome businesses.

4:33

Warren called that decision a slap on

4:35

the wrist. She and other lawmakers argue

4:38

that no single company should control so

4:41

much of the internet.

4:42

And Monday's outage proved their point.

4:45

It highlights the scale of dependence on

4:48

a few major players. One error, one

4:51

company, and almost the entire web goes

4:54

offline. For millions, it was a reminder

4:58

the internet may feel endless, but it

5:01

still runs through a few very powerful

5:04

hands.

5:07

Want the facts,

5:08

the latest developments,

5:10

news that gets straight to the point?

5:12

Well, we've got all three just for you.

5:15

This is First Post Live, a brand new

5:17

show, your window into what really

5:19

matters.

5:20

Don't miss it.

0:00

The internet just died yesterday with so

0:02

many websites being down for more than

0:04

12 hours. We're talking about Amazon,

0:06

Snapchat, Google, Hulu, Netflix, Reddit,

0:09

Lyft, Spotify, Starbucks, T-Mobile,

0:11

Verizon, Roblox, Fortnite, a bunch of

0:13

airlines, Nintendo Switch Online, Venmo,

0:16

and Zoom. And then a bunch of Amazon

0:18

apps like Alexa, Kindle, Ring, Prime

0:20

Video. That's a lot. That's like a huge

0:21

chunk of the internet that most people

0:23

use. So, what exactly happened? We're

0:25

going to look into it right here, cuz

0:26

this is actually crazy. And this comes

0:28

after YouTube itself went down like last

0:30

week sometime. So what's going on? Why

0:32

are so many websites going down now? So

0:34

apparently it was due to a DNS error,

0:36

whatever the heck that means. So

0:38

Amazon's web service went down, AWS, and

0:42

a lot of different platforms and

0:43

websites actually use AWS to just have

0:46

things running on their websites. AWS is

0:48

a cloud computing infrastructure. It

0:50

powers a lot of the internet. It's like

0:52

it it backs so many platforms and

0:54

websites. Like all those things we said,

0:55

just one example, Snapchat, all of their

0:58

stuff is backed on Amazon Web Services.

1:00

So if they go down, if Amazon goes down,

1:02

Snapchat and all these other things go

1:04

down. It's kind of crazy that Google

1:05

went down. Like you think Google and

1:07

Amazon would separate, but I guess not.

1:09

So I just checked and Google doesn't use

1:10

Amazon Web Services, but but they do for

1:12

maybe specific products and purposes,

1:14

but it's still really weird that Google

1:15

and Amazon and YouTube all went down in

1:18

like the last week. Anyways, this all

1:20

first happened at 3:00 a.m. Eastern. So,

1:22

it was in the middle of the night that

1:24

this happened. Down detector was going

1:26

crazy. Like, look at all of these

1:27

different websites that went down.

1:29

There's so many. Oh my god. Even just a

1:32

bunch of video games. Adobe, Delta, like

1:34

all these airlines. Oh man, that's

1:36

actually insane. And then Amazon

1:38

responded at 6:35 a.m. talking about a

1:40

partial fix. And then 6 hours later at

1:44

6:00 p.m. there was a full restoration.

1:46

Like, they fully fixed it. They fully

1:48

fixed whatever happened. The outage was

1:50

from 3:00 a.m. to 6:00 p.m. That's over

1:52

12 hours. So, Amazon responded to what

1:54

happened and their response is just kind

1:55

of weird and vague. Amazon said that the

1:57

outage was likely caused by issues

1:59

related to its DNS. Likely they do they

2:02

even know fully likely. What the heck?

2:05

And DNS is their domain name system

2:07

which converts website addresses into

2:09

numeric ones. And it people describe it

2:11

as a phone book for the internet. So,

2:12

it's literally just how things work.

2:15

It's how websites work and process

2:17

information and handle stuff. It's also

2:19

described as the plumbing for online

2:20

connectivity. So, if that goes down,

2:23

you're kind of messed up for a long time

2:24

and that is what happened. It was most

2:26

of the day yesterday that all these

2:28

services that use Amazon was just

2:29

completely down. There were also other

2:31

weird cases like a lot of smart home

2:33

tech was using these Amazon web

2:35

services. So, people with smart

2:36

mattresses found that their mattresses

2:38

were stuck in the morning in like an

2:40

incline position or overheating cuz they

2:42

had the mattress heated. Like, that's

2:44

definitely not good. And I don't know

2:45

why they need internet to have all those

2:47

features. just make it in the device.

2:49

It's a mattress. I'm pointing here

2:51

because my bed's right here. It does

2:53

spark concerns with all of these

2:54

services using Amazon Web Services. I

2:57

think some websites need to consider

2:58

maybe diversifying because if you all

3:00

rely on one thing, it's obvious that if

3:03

it goes down, a lot of the internet goes

3:04

down as well. But yeah, YouTube went

3:06

down just last week and now this

3:08

happening and and Google at the same

3:09

time reporting outages. It's just really

3:11

like concerning. Like why are all these

3:13

websites going down lately? Is something

3:15

else going on? It makes you wonder, but

3:17

like it could also just be coincidence.