What You Can Learn From CrowdStrike Breaking The World

preview_player
Показать описание
CrowdStrike recently took down nearly all major companies across the globe. Airlines, banks, hospitals, police stations, and many more large scale companies were brought down entirely. In this video I will explain what happened, why it caused such massive problems, and what you can learn from this massive problem to make your company more resilient to potential issues.

🌎 Find Me Here:

⏱️ Timestamps:

00:00 - Introduction
00:45 - What Is CrowdStrike
01:35 - Why It Broke The Internet
03:29 - What You Can Learn From This #1
07:26 - What You Can Learn From This #2
09:40 - What You Can Learn From This #3

#CrowdStrike #WDS #WebDevelopment
Рекомендации по теме
Комментарии
Автор

Crowdstike’s CEO, George Kurtz, was also CTO of McAfee back in 2010 when McAfee sent an update that also caused global crashes for Windows XP machines. Let that sink in.

RubixCubed
Автор

0:03 i love this shot ;D Magazines usage 10/10

MaxSharpy
Автор

What an occasion in a vacation.
Happy vacation Kyle.

풍월상신
Автор

You helped me survive in my first internship, and how I have 3 + years of experience.
Respect ❤

CodeWithMuhammadAhsan
Автор

Apparently they rolled out a change in April that shutdown a number of Linux servers. It wasn’t noticed as much because most Linux servers don’t run crowd strike, and those that do don’t often auto install updates.

steves
Автор

When we are talking about NPM this logic make total sense. 3:36
When we are talking about security you need the last not breaking version always.

If you want to mix both NPM and security, you should have as less dependencies as possible. It can be sound crazy, but less lines of code means less bugs and less non used feature that can help in an attack.

diogotrindade
Автор

Using a few versions older might pose security threats.
When updates are pushed to 3rd party libraries, or to software products, there is usually a detailed report of what bugs or security risks were fixed.
Then hackers can just check out the security risks and try those on many machines until they hit a jackpot.

erickrief
Автор

Uncle Bob mentions this being a potential on everyone of his clean architecture talks.

ZwCode
Автор

Now imagine Cloudflare doing something like Crowdstrike. It will actually ruin half of the Internet. Since it's Windows who used Crowdstrike, only windows using businesses were affected. But cloudflare could easily break DNS of half of the intenet. Or just disable access to it.
Monopoly? Never heard of it.

BouhartsevStudio
Автор

Will a simple e2e testing not resolve this instead of maintaining library version?

uchennahenry
Автор

Can you make a video on project deploying using domain?

akapain
Автор

Imagine using Windows for mission critical tasks. LTS Unix/Linux is where it is at. Windows has not had the basic respect for it's users in year for the autonomy of their machines. Hence the forced restart update "feature" pushed out in windows 8/10 for personal licenses. I wish we lived in a world where airlines had invested in dedicated branches to unix/linux back 25 years ago to run on. Windows servers is one of the most crazy things that were a fad for sometime. Imagine all the bloatware and sketchy proprietary code on your desktop, now poorly ported to run as a server which handled critical tasks like running an airline.

WYO_Dirtbag
Автор

So what you're saying is this is about to become a Rust channel so we don't get anymore null ptr errors. It compiles to WASM, so there's some WebDev content right there!

TylerR
Автор

What can we learn from this? 0:39
CrowdStrike runs in kerner mode and that is a huge mistake. macOS, iOS and others have APIs to this type of interactions.
Essencial companies cannot stop working and this situation proff that we cannot have boot-start drives running on kernel mode. It is not safe when a mistake happen and creates a huge point of failure. So I believe that companies need to use better OS that are better build for failure without any nonsenses running on the kernel.

diogotrindade
Автор

Glad I'm not whoever is responsible of that fuck up nor whoever who's supposed to fix it.

SomaStroke
Автор

this was a kernel level update, iirc. insane for this to make it through certification "by accident".

tjk_
Автор

What I am wondering is; with the mentality of "letting other people test the newest version", can there come a situation where end user are more resistant to applying newer updates if their current version is stable enough? In other words: who are the other people?

ManjiVegeta
Автор

actually what we learned from this is, that companies can stop your device no matter where are you or when.

scienctv
Автор

I still find it funny that a security company chose a name that sounds like malware.

stephenriggs
Автор

Don't know if it is a good idea to run older versions of virus definitions and rules

KraasRas