Business Global IT outage live updates: Australian banks, airlines, media outlets taken offline - Bloody Bitch Bastard

There's a global outage of MicroSoft Windows machines currently, amusingly. This website is laid out in a very frustrating way, so I've included the main excerpts here:
https://www.abc.net.au/news/2024-07...-microsoft-banks-airlines-australia/104119960 (archive)

There are reports of IT outages affecting major institutions in Australia and internationally.
The ABC is experiencing a major network outage, along with several other media outlets.
Crowd-sourced website Downdetector is listing outages for Foxtel, National Australia Bank and Bendigo Bank.

Like a number of other organisations, global issues affecting CrowdStrike and Microsoft are disrupting some of our systems.
The issue is causing some holdups for some of our customers and we thank them for their patience.
There is no impact to our fixed or mobile network which continue to operate.
Calls to our Triple Zero contact centres are not affected, but we understand some state emergency services are also impacted and we are working with them to implement backup processes.

CrowdStrike ran a recorded phone message on Friday saying it was aware of reports of crashes on Microsoft's Windows operating system relating to its Falcon sensor.
"Thanks for contacting CrowdStrike support. CrowdStrike is aware of reports of crashes on Windows … related to the Falcon sensor," a prerecorded message played when a Reuters reporter called the company's technical support.

University of Melbourne lecturer in cyber security Shaanan Cohney says there appear to be two separate things happening at once to cause the mass outages we are seeing.
The first issue, he says, appears to have been caused by a piece of software developed by a company called CrowdStrike.
"It's a computer security vendor that provides a monitoring service to large enterprises so they can see on computers within their control if there's any indications of suspicious activity or things that would require a security alert or to lock down the computer," Dr Cohney says.
"Because this software needs to see everything that is going on, it's very tightly integrated into the computer's software, so when you install it, it asks for a lot of permissions so that it can ask for everything going on on the computer.
"However, because it's in such a privileged position, if something goes wrong with it, if there's a programming mistake it has the capability to bring down the entire computer.
"If someone makes the wrong type of mistake it can bring the whole system down.
"As far as we can tell what it looks like happened with this piece of software is the company issued a significant update and something in the update went wrong.
"Engineers at the company and those outside are scrambling to try to pinpoint the source so they can try to pinpoint the problem so that's why companies are telling their employees to shut down their computer in order to prevent them from updating so those employees can maintain some minimal capabilities and have access to documents that are offline."
Reporting by Andi Yu

lcimg-7620c845-53fb-4caa-84ce-f61f76fa0ee6.jpeg
lcimg-5416bf03-1f8b-4709-b65b-ab1615785cba.jpeg
 
A colleague told me some shit linked to medical imaging cloud storage went bust and they're not able to do anything remotely.
I wonder how many people died from Microsoft's shitty code today. Imagine getting scheduled to some surgery just for the people to say "welp sorry, our system is more dead than you are, come back tomorrow".

Sucks to be Boogie2988. If this happened when he was doing those streams, he could have legitimately lied about the portals being down due to this fiasco. It'll be hard to prove otherwise.
Confirmed. Every computer in the hospital is down.
 
Not defending Microcock here, but every machine not running Crowdstrike is unaffected. The crux of the issue is that a third-party software is fucking with the kernel (it's cybersec, so it's granted) and has a good chunk of the market share. The market share for Crowdstrike was 18.5% in Q2 2023. (archive)
The second fiscal quarter of 2024 recently ended, so I'm just rounding the numbers here to about ~20%. That means a fifth of the global Windows systems, which use this software are currently fucked and the damages are astronomical already. There are workarounds to fix this issue, but it's already too late. I don't think that Crowdstrike will survive this disaster in the long run.

Now imagine Riot fucking up Vanguard (kernel level anti-cheat rootkit software) at this magnitude. It would be funny, not horrifying like this case.
Well that 20% is about to go to some other company in the future because fucking hell this is one way to tank trust.
 
Same shit on AWS EC2 and GCP because for some stupid reason none of these big public clouds give you actual graphical console access to the server, just serial port. If you have EMS and SAC enabled you may be able to recover via safe mode but I haven't tried this and I know Windows Recovery does not work via serial based on prior experience.
Supposedly if you can get to the Windows recovery screen, all you need to do is rename the CrowdStrike folder in
Kod:
C:\Windows\System32\drivers
to break the agent. Then you will be able to reboot.
 
Yet another thread where people bleat replies without reading the article. These security agents rarely work right on Mac and the nebula of Linux OSes by default so obviously Crowdstrike Falcon is going to break Windows. Journoscum sensationalize it as "oohh Windows broke!!!" and since it's Microshit being spat on the average KFer eats it right up.
Of all the topics this shit got you fired up?

I hope you have a good week, man, damn.
 
Supposedly, I read that someone's smart fridge isn't working because of this. It could be a joke/meme referencing that water dispenser update from a few years back.
Imagine buying a wifi-connected IOT fucking fridge.
In this case, they deserved it. Enjoy rotten food.
Whatever backend it uses could be down from this. But who the fuck honestly buy a smart fridge? The fuck is wrong with people. People are so fucking dumb now they need every device to be a smart device. Hurr durr i don ned be smrt. my stuf smrt 4 me.
 
Supposedly, I read that someone's smart fridge isn't working because of this. It could be a joke/meme referencing that water dispenser update from a few years back.
Imagine buying a wifi-connected IOT fucking fridge.
In this case, they deserved it. Enjoy rotten food.
I fucking hate the computerization of everything, and the need to throw wi-fi and Bluetooth on them. It's a fucking fridge, it doesn't need microchips and all that shit. Knock it the fuck off.

Whatever backend it uses could be down from this. But who the fuck honestly buy a smart fridge? The fuck is wrong with people. People are so fucking dumb now they need every device to be a smart device.
Watching what the iPhone did to people; the small convenience it offers outweighs all the potential problems, because these people dare not to try and gaze at the horizon. The people who do bring this shit up; conspiracy theorists or lame nerds who can't get with the times. Because they don't know most of this shit is assembled by 8 year old Chinese children and coded by sex starved useless code camp Pajeets.
 
Of all the topics this shit got you fired up?

I hope you have a good week, man, damn.
In fairness, he's really feeling it this morning.
Supposedly, I read that someone's smart fridge isn't working because of this. It could be a joke/meme referencing that water dispenser update from a few years back.
Imagine buying a wifi-connected IOT fucking fridge.
In this case, they deserved it. Enjoy rotten food.
Unless that fridge is running Windows and CrowdStrike, that issue is unrelated to this outage.
 
How in the hell did this happen? Did they push out an update without testing it at all? A bug bringing down systems at this scale should have been caught by any halfway decent regression test suite.
 
The simplest summary I can understand of this so far:

Corporate cyber security software "CrowdStrike Falcon Sensor" pushed a faulty update to all customers running the software on Windows platforms, making them unable to boot
The company already stopped pushing the update, but as many systems using the software are servers that run 24/7, a huge chunk of damage from downtime of major infrastructure has already been inflicted and is still on-going
The fix for affected machines is relatively simple (boot into safe mode and delete the folder containing the update), but will have to be done manually for every machine; which is a huge pain for corporations running thousands of computers/servers, even more so for systems normally accessed remotely ("on the cloud").

How in the hell did this happen? Did they push out an update without testing it at all? A bug bringing down systems at this scale should have been caught by any halfway decent regression test suite.
It sounds like they compiled a new build and pressed the "push to consumers" button instead of the "push to test suite" button.
Probably the fault of pajeets tbh
 
Nigger CEO hoping people are unable to parse IT speak. A fucky wucky in the shit you push is a security issue; I don't care how you want to try and dance around the issue, if you push shit into production and you take out this much infrastructure, you have a fucking problem. Maybe this is just me, but trust is a security issue; even if this is a fucking oopsie by one of your Pajeets and not a glow op, no one should trust you to do shit right. Because okay, you fucking take out half the world with this update; when can we expect one of your retarded code monkeys to put ransomware into your product?
I have 100% confidence in Cloudstrike.

t. Alejandro Mayorkas
 
The fix for affected machines is relatively simple (boot into safe mode and delete the folder containing the update), but will have to be done manually for every machine; which is a huge pain for corporations running thousands of computers/servers, even more so for systems normally accessed remotely ("on the cloud").
The annoyance is actually the other way around: servers and cloud systems are fixable in bulk because you can remote access them and just cook up a ghetto script to automate it, but end user workstations have to be done manually one by one because there's no way to remotely access them in safe mode and if BitLocker is enabled (it often is), you also have to fish out the 48 character recovery key and punch it in before you can even attempt the fix. A large site with 10k+ workstations is going to be EXCRUCIATING.
 
My man just arrived home from work, all of the computers are down, so nobody can get anything done.
Didn't even make it to lunchtime.
Still, no-one is going to complain about a surprise half-day off on a Friday. POETS (piss off early, tomorrow's Saturday) day is the perfect day for this.
This is in Glasgow,
btw.
 
Wstecz
Top Na dole