This isn’t a gloat post. In fact, I was completely oblivious to this massive outage until I tried to check my bank balance and it wouldn’t log in.
Apparently Visa Paywave, banks, some TV networks, EFTPOS, etc. have gone down. Flights have had to be cancelled as some airlines systems have also gone down. Gas stations and public transport systems inoperable. As well as numerous Windows systems and Microsoft services affected. (At least according to one of my local MSMs.)
Seems insane to me that one company’s messed up update could cause so much global disruption and so many systems gone down :/ This is exactly why centralisation of services and large corporations gobbling up smaller companies and becoming behemoth services is so dangerous.
We’re all going to be so smug.
Me too. Additionally, I use guix so if a system update ever broke my machine I can just rollback to a prior system version (either via the command line or grub menu).
That’s assuming grub doesn’t get broken in the update…
True, then I’d be screwed. But, because my system config is declared in a single file (plus a file for channels) i could re-install my system and be back in business relatively quickly. There’s also guix home but I haven’t had a chance to try that.
I would definitely recommend using guix home because having a seperate config for you more user facing stuff is so convenient (plus no need for root access to install a package declaratively) (side note take this with a grain of salt because I don’t use gnu guix I use nixos)
Most people are completely oblivious because it only affects people using crowdstrike, which practically excludes general consumers.
I just had an Amazon package delayed for a week it says. It doesn’t name names but…
A small number of deliveries may arrive a day later than anticipated due to a third-party technology outage.
Microsoft should test all its products on its own computers, not on ours. Made an update, tested it and only then posted it online.
Microsoft has nothing to do with this. This is entirely on Crowdstrike.
Crowdstrike already killed some Linux machines. Let’s not pretend Windows is at fault here or Linux is magically better in this area. No one is immune from software that can run as a kernel module going bad.
But, my superiority!
Every system has its faults. And I’m still going to dogpile the system with the most faults. But hell Microsoft did buy GitHub, Halo, MineCraft, and a million other things they will probably find a way to buy Linux and ruin it for us just like they ruin everything else.
Let’s see, …we are somewhere in between Extend and Extinguish on the roadmap.
Edit: Case & Point, RIP RedHat & IBM and GitHub CoPilot, what a great idea. RIP Atom Editor and probably a million other things. Do we have a KilledByMicrosoft website yet? I hope people in the pharmacy could get their prescriptions or we might have to add peoples names to the list.
None of this has to do with the current outage though.
I hope people in the pharmacy could get their prescriptions or we might have to add peoples names to the list.
Which isn’t Microsoft’s fault. Linux systems have also been taken down by Crowdstrike’s fuck ups in the recent past.
Microsoft has many faults and I’ll criticize them as I please. And if Linux is a culprit in a global outage someday I’ll contemplate criticizing them too.
This “Not Microsoft’s Fault” comes off as white knighting for Muh Billion Dolla Corporation.
Do we really need to SIMP for the company town.
Microsoft, Google, Apple, Amazon and others deserve every ounce of vitrol they earn through their shitty practices. Again I am criticizing them for being shitty not for the particulars of System X vs System Z but for the aftermath.
Except they haven’t done anything shitty this time. What you are doing would be a bit like claiming the Nazis are responsible for micro plastics. Like yeah Nazis are shit but making false allegations is just giving their defenders something to throw in your face. It makes you, and everyone who is critical of Microsoft look dumb. How about you criticize the company that actually screwed up? They are also a multi-billion dollar company, yet you aren’t blaming them for something that is clearly their fault.
Sure you can criticize as much as you want but if you are wrong in your criticism it just damages all of your criticism over all.
In my opinion it is important to state facts not fiction. This was not Microsoft’s fault, no matter how much you hate Microsoft it still wasn’t there fault and saying that is was is incorrect and doesn’t solve the issue.
Well said, that’s one of the points I have been trying to get across.
I get where you are coming from, but this event is pretty much entirely the fault of Crowdstrike and the countless organizations that trusted them. It’s definitely a show of how massive outages are more likely when things are overly centralized and proprietary, and managed by big, shitty, profit driven organizations. Since crowdstrike operates in kernel space, it doesn’t matter which operating system it’s on, it can break it if it does something stupid. In fact they managed to break some redhat machines not too long ago, and some Debian machines not long before that. It’s just the impact wasn’t as far reaching as this recent utter fuckup, just because fewer critical machines were affected, so we didn’t hear about those smaller fuckups in the news.
Yes, thank you, exactly. The centralized model has its benefits but it also can act as a single point of failure.
If I was going to analyze from an engineering perspective I would focus on when these inevitable events occur due to human error do we have adequate tools to roll back updates? Do we snapshot OS drives before updates? Is there adequate Safe Mode or Fallback Tools to diagnose which files are offending in order to allow the user to remove them.
In my view the windows user isn’t dignified to have the skills or intelligence needed to workaround a “setback” issue like the one yesterday.
It doesn’t help that NTFS is missing modern capabilities, or that there isn’t easy to use DIFF for the layman to understand which files were added to the filesystem that may be causing the breakage.
To be fair though even with those pot holes filled the entire design paradigm of Windows and a proprietary platform is part of the problem. Software is not broken up into package modules that can be assembled into a functioning system it is encumbered with “anti-piracy” boogie man where the software treats the user as an enemy and is designed to break.
Linux isn’t like that. I’ve cloned many distro drives and swapped them into new machines and with 1 or 2 tweaks they JustWork
I see many people on the net defending Microsoft as blameless for technical reasons.
My criticisms were that Microsoft just sucks as you interpreted correctly and offered a eloquent summary. Thank You.
Where I think the entire conversation should move is –
What are the design flaws that allowed this to happen?
“More Rust & Less C” I see some people suggest as this was allegedly a null pointer issue.
And is Windows Broken By Design? My opinion answer - Yes.
(Okay, and what to do about it before the next billion dollars is lost. I would think critical infrastructure should have a model similar to NixOS in immutability but that’s just my opinion.)
Windows does have a fallback mode called safe mode and that’s exactly what’s being used to fix this utter mess.
Package management isn’t going to save you from this as it didn’t save the Linux systems affected last time. It didn’t stop Arch Linux from failing to boot after a Grub update either.
Windows also has drive cloning tools, that isn’t unique to Linux.
NixOS isn’t immutable. It’s not an a/b root system and / isn’t read only. Rather it’s what’s known as reproducible. I am not convinced NixOS would make this any easier either given how simple the fix was. Funnily enough though tools exist called ansible and puppet for configuring systems in repeatable ways that apply to both other Linux systems, Windows systems, and even macOS.
There are like one or two valid points in this whole comment and the rest is pretty much falsehoods and misconceptions.
Edit: Forgot to mention tools exist to make Windows immutable as well. So that is an option.
Windows does have a fallback mode called safe mode and that’s exactly what’s being used to fix this utter mess.
The other fix was reboot your Windows computer at least 15 times.
Package management isn’t going to save you from this as it didn’t save the Linux systems affected last time. It didn’t stop Arch Linux from failing to boot after a Grub update either.
Not everyone was affected though :
How come not everyone was impacted?
Prior to the most recent version, grub only registered the fwsetup if detected support. If your machine detected support, you would have had the fwsetup command registered and the failure wouldn’t occur.
Also fyi Red Hat and IBM are still around and aren’t really a force for good anyway. Stop SIMPing for large companies.
Hilarious. I am sure that, out of principle, you have stopped using all the software that Red Hat contributes to your distribution.
If it is ok with you, I am not going to define my morality in terms of corporate interest. They are not my friends but I do not believe that shutting on their contributions does much for me either.
I am not shitting on their contributions. All I am saying is that as a large company they aren’t anymore my friend than Microsoft. Generally they still exist and make contributions. Microsoft didn’t kill them like the person I am replying to is insinuating.
Even 911 is impacted
That’s potentially life threatening. I wonder if 112 in other countries is affected, it shouldn’t be but at this point I’m afraid it is.
In the Netherlands 112 is fine, most critical systems are. It’s mostly airports that are getting fucked by this it seems.
Banks and PSPs are fine here too.
deleted by creator
It’s proving that POSIX architecture is necessary even if it requires additional computer literacy on the part of users and admins.
The risk of hacking (which is what Crowdstrike essentially does to get so deeply embedded and be so effective at endpoint protection) a monolithic system like Windows OS is if you screw up the whole thing comes tumbling down.
deleted by creator
deleted by creator
deleted by creator
deleted by creator
deleted by creator
deleted by creator
Agreed.
I’ve just spent the past 6 hours booting into safe mode and deleting crowd strike files on servers.
Feel you there. 4 hours here. All of them cloud instances whereby getting acces to the actual console isn’t as easy as it should be, and trying to hit F8 to get the menu to get into safe mode can take a very long time.
Ha! Yes. Same issue. Clicking Reset in vSphere and then quickly switching tabs to hold down F8 has been a ball ache to say the least!
Just go into settings and add a boot delay, then set it back when you’re done.
What I usually do is set next boot to BIOS so I have time to get into the console and do whatever.
Also instead of using a browser, I prefer to connect vmware Workstation to vCenter so all the consoles insta open in their own tabs in the workspace.
Can’t you automate it?
Since it has to happen in windows safe mode it seems to be very hard to automate the process. I haven’t seen a solution yet.
Sadly not. Windows doesn’t boot. You can boot it into safe mode with networking, at which point maybe with anaible we could login to delete the file but since it’s still manual work to get windows into safe mode there’s not much point
It is theoretically automatable, but on bare metal it requires having hardware that’s not normally just sitting in every data centre, so it would still require someone to go and plug something into each machine.
On VMs it’s more feasible, but on those VMs most people are probably just mounting the disk images and deleting the bad file to begin with.
I guess it depends on numbers too. We had 200 to work on. If you’re talking hundreds more than looking at automation would be a better solution. In our scenario it was just easier to throw engineers at it. I honestly thought at first this was my weekend gone but we got through them easily in the end.
The real problem with VM setups is that the host system might have crashed too
I work in hospitality and our systems are completely down. No POS, no card processing, no reservations, we’re completely f’ked.
Our only saving grace is the fact that we are in a remote location and we have power outages frequently. So operating without a POS is semi-normal for us.
I’ve worked with POS systems my whole career and I still can’t help think Piece Of Shit whenever I see it
I am born too late to understand what Y2K problem was, this (the result) might be what people thought could happen.
Yep pretty much but on a larger scale.
1st please do not believe the bull that there was no problem. Many folks like me were paid to fix it before it was an issue. So other than a few companies, few saw the result, not because it did not exist. But because we were warned. People make jokes about the over panic. But if that had not happened, it would hav been years to fix, not days. Because without the panic, most corporations would have ignored it. Honestly, the panic scared shareholders. So boards of directors had to get experts to confirm the systems were compliant. And so much dependent crap was found running it was insane.
But the exaggerations of planes falling out of the sky etc. Was also bull. Most systems would have failed but BSOD would be rare, but code would crash and some works with errors shutting it down cleanly, some undiscovered until a short while later. As accounting or other errors showed up.
As other have said. The issue was that since the 1960s, computers were set up to treat years as 2 digits. So had no expectation to handle 2000 other than assume it was 1900. While from the early 90s most systems were built with ways to adapt to it. Not all were, as many were only developing top layer stuff. And many libraries etc had not been checked for this issue. Huge amounts of the infra of the world’s IT ran on legacy systems. Especially in the financial sector where I worked at the time.
The internet was a fairly new thing. So often stuff had been running for decades with no one needing to change it. Or having any real knowledge of how it was coded. So folks like me were forced to hunt through code or often replace systems that were badly documented or more often not at all.
A lot of modern software development practices grew out of discovering what a fucking mess can grow if people accept an “if it ain’t broke, don’t touch it” mentality.
Was there patching systems and testing they survived the rollover months before it happened.
One software managed the rollover. But failed the year after. They had quickly coded in an explicit exception for 00. But then promptly forgot to fix it properly!.
Kinda I guess. It was about clocks rolling over from 1999 to 2000 and causing a buffer overflow that would supposedly crash all systems everywhere causing the country to come to a hault.
Most old systems used two digits for years. The year would go from 99 to 0. Any software doing a date comparison will get a garbage result. If a task needs to be run every 5 minutes, what will the software do if that task was last run 99 years from now? It will not work properly.
Governments and businesses spent lots of money and time patching critical systems to handle the date change. The media made a circus out of it, but when the year rolled over, everything was fine.
Also a lot of people were “on call” to handle any problems when the year changed, so the few problem that had passed unnoticed when doing the fixed and did pop up when the year changed, got solved a lot faster than they normally would.
We also got the worst version of Windows ever, ME. Tho maybe with all the BS they’ve done with 11 that might change.
Millennium Editions ruin everything!! 🤬
I’d use ME before the adware that is the current version. It wasn’t that bad, it was just Win98 with some visual slop on top that crashed slightly more often.
I’m not sure I’d stick to calling it the worst version “ever” since MS is trying really hard to out do themselves.
One program I tested went from (31,12,99) to (01,01,100). Its front end formatted the date and added the century, so it showed 1 January 2000 as 01/01/19100
That wasn’t fixed. The fault didn’t affect processing (the years were wrong but had the correct offset between them) and was only visible to internal users, and also that system was expected to be retired in 2004
And it was okay because a lot of people worked really really hard to make it be okay.
Y2K was going to be the end of civilisation. This was basically done by the time I woke up today.
after reading all the comments I still have no idea what the hell crowdstrike is
Company offering new-age antivirus solutions, which is to say that instead of being mostly signature-based, it tries to look at application behavior instead. If Word was exploited because some user opened not_a_virus_please_open.docx from their spam folder, Word might be exploited and end up running some malware that tries to encrypt the entire drive. It’s supposed to sniff out that 1. Word normally opens and saves like one document at a time and 2. some unknown program is being overly active. And so it should stop that and ring some very loud alarm bells at the IT department.
Basically they doubled down on the heuristics-based detection and by that, they claim to be able to recognize and stop all kinds of new malware that they haven’t seen yet. My experience is that they’re always the outlier on the top-end of false positives in business AV tests (eg AV-Comparatives Q2 2024) and their advantage has mostly disappeared since every AV has implemented that kind of behavior-based detection nowadays.
AV, EDP they offer other solutions as well. I think their main selling point is tamper-proof protection as well.
Seems to be some sort of kernel-embedded threat detection system. Which is why it was able to easily fuck the OS. It was running in the most trusted space.
That’s hell of a strike to the crowd
Crowdstrikeout
Crowd II: The Struckening
US and UK flights are grounded because of the issue, banks, media and some businesses not fully functioning. Likely we’ll see more effects as the day goes on.
I love how everyone understands the issue wrong. It’s not about being on Windows or Linux. It’s about the ecosystem that is common place and people are used to on Windows or Linux. On windows it’s accepted that every stupid anticheat can drop its filthy paws into ring 0 and normies don’t mind. Linux has a fostered a less clueless community, but ultimately it’s a reminder to keep vigilant and strive for pure and well documented open source with the correct permissions.
BSODs won’t come from userspace software
Crowdstrike does have linux and mac version. Not sure who runs it
That’s precisely why I didn’t blame windows in my post, but the windows-consumer mentality of “yeah install with privileges, shove genshin impact into ring 0 why not”
Linux can have the same issue. We have to keep the culture on our side here vigilant and pure near the kernel.
I deployed it for my last employer on our linux environment. My buddies who still work there said Linux was fine while they had to help the windows Admins fix their hosts.
While that is true, it makes sense for antivirus/edr software to run in kernelspace. This is a fuck-up of a giant company that sells very expensive software. I wholeheartedly agree with your sentiment, but I mostly see this as a cautionary tale against putting excessive trust and power in the hands of one organization/company.
Imagine if this was actually malicious instead of the product of incompetence, and the update instead ran ransomware.
If it was malicious it wouldn’t have had the reach a trusted platform would. That is what made the xz exploit so scary was the reach and the malicious attempt.
I like open source software but that’s one big benefit of proprietary software. Not all proprietary software is bad. We should recognize the ones doing their best to avoid anti consumer practices and genuinely try to serve their customers needs to the best of their abilities.
I would be too, except Firefox just started crashing on Wayland all the morning D;
New Nvidia driver?
Yes but I upgraded to 555 at least a week or two ago and it started crashing a couple of days ago, I think there’s an issue with explicit sync
explicit sync is used, but no acquire point is set
If you Google this you’ll find various bug reports
