https://www.neowin.net/news/crowdstrike-broke-debian-and-rocky-linux-months-ago-but-no-one-noticed/
CrowdStrike recently caused a widespread Blue Screen of Death (BSOD) issue on Windows PCs, disrupting various sectors. However, this was not an isolated incident, CrowdStrike affected Linux PCs also.
To avoid such issues in the future, CrowdStrike should prioritize rigorous testing across all supported configurations.
Bold of them to assume there's a future after a gazillion off incoming lawsuits.
I was listening to a podcast earlier, and they mentioned the fact that their legal liability may, in fact, be limited because of specific wording in most of their contracts.
In other words, they may actually get away with this in the short term. In the long-term, however, a lot of organizations and governments that were hit by this will be reevaluating their reliance on such monolithic tech solutions as crowdstrike, and even Microsoft.
So you may be right, but not for the reasons you think.
and even Microsoft
(x) doubt
They had decades to consider Microsoft a liability. Why start doing something about it now?
Because cybersecurity is becoming more of a priority. The US government has really put their attention on it in the last few years.
I was in IT back in 2001 when the Code Red virus hit. It was a very similar situation where entire enterprises in totally unrelated fields were brought down. So many infected machines were still trying to replicate that corporate networks and Internet backbone routers were getting absolutely crushed.
Prior to that, trying to get real funding for securing networks was almost impossible. Suddenly security was the hottest topic in IT and corporations were throwing money at all the snake oil Silicon Valley could produce.
That lasted for a couple years, then things started going back to business as usual. Microsoft in particular was making all sorts of promises and boasts about how they made security their top priority, but that never really happened. Security remained something slapped on at the end of product development and was never allowed to interfere with producing products demanded by marketing with inherently insecure designs.
You're absolutely right. Everyone will be very worried and talk about the importance of security in the enterprise and yada yada yada until a cool new AI spreadsheet software comes out and everybody forgets to even check if their firewall is turned on.
But with that being said, if you have been looking for a good time to ask for cybersecuity funding at your org, see if you can't lock down 5 years worth of budget while everyone is aware of the risk to their businesses.
Contracts aren’t set in stone. Not only are those contracts modified before they are accepted by both parties, it’s difficult to limit liability when negligence is involved. CS is at worst going to be defending against those, at best defending against people dumping them ahead of schedule against their contracted term length.
Oh so you can fire QA department, get absolutely destructive update to millions of systems across the globe and this gross negligence doesn't matter because of magic words in a contract? I don't think so.
Then how else is their legal liability is limited?
They killed off their QA department to chase profits which resulted in a broken product that crippled hundreds of organizations across the globe.
They don't get to just shrug, say oopsie, and point at the contract.
They mean after Crowdstrike gets sold, the new company promises a more rigorous QA, and quietly rebrands it.
I think you mean after they sell their assets to a new company. Leave the lawsuits with the old company who will shut down.
Additionally, organizations should approach CrowdStrike updates with caution
We would if we were able to control their "deployable content".
I read on another thread that an admin was emulating a testing environment by blocking CrowdStrike IPs on their firewall for the whole network before each update, with the exception of a couple machines. It's stupid that he has to do this but hey, his network was unaffected
Serious question, can you not? There isn't an option to...like...set a review system first?
Turns out it was a content update that caused the driver to crash but the update itself wasn't a driver (as per their latest update.)
Found this post that explains what happened in detail: https://lemmy.ohaa.xyz/post/3522666
As an application developer (rather than someone who can/does code operating systems) I was just left open-mouthed …
Looks like they’re delivering “code as content” to get around the rigour of getting an updated driver authorised by MS. I realise they can’t wait too long for driver approval for antivirus releases but surely - surely - you have an ironclad QA process if you’re playing with fire like this.
Setting the update policy to N-2 (or any other configuration) would not have avoided the issue. The Falcon sensor itself wasn’t updated, which is what the update policy controls. As it turns out, you cannot control the content channel updates - you simply always get the updates.
We would if we were able to control their “deployable content”.
Minimum safe distance.
not sure if you're being sarcastic, but if anything this news paints linux deployment in an even better light.
Nah, but there were some Linux evangelists claiming this couldn't possibly happen to Linux and it only happened to Windows because Windows is bad. And it was your own fault for getting this BSOD if you're still running Windows.
And sure, Windows bad and all, but this one wasn't really Microsofts fault.
The sane ones of us know well that a faulty driver is a faulty driver, but! Linux culture is different. Which is why this happened so spectacularly with Windows. EDIT: and not with Linux
I've had the proprietary Nvidia driver crash my whole system a few times. Hoping their new open-source driver (not nouveau, I mean the new out-of-tree open-source one) is better.
I had X crash due to Nvidia under FreeBSD a few times, and fewer kernel panics due to it. Never used Linux with Nvidia though.
Also, even if they do, you can choose to not load it.
It amused me that so many people had this installed, but had no idea what it was for.
if they dont know the boot sequence is a thing maybe their opinion on this doesnt really matter 🤷🏼
I'm not shocked at all, but there seems to be a very sizable number of people on Lemmy who think if people just used Linux there'd never be another problem or exploit again, which is ridiculous. Mac users used to feel the same way until the market share started to grow and all of the sudden you're seeing news of serious exploits.
Haven't you heard 4% market is captured by Linux , it's the ONLY saviour os out there , windows users and macos users are idiots and all Lemmy Linux dudebros grandpa's are using Linux without single problem. Despite the fact that each Linux had it's own shell and there is no escape from terminal ( in 2024) if you even as try to use something more complicated. ;)
For almost every use case a normal user needs, there is a gui. You do not need the terminal.
Tell me where to find executables for programs installed without using Terminal , a very very clickable task in windows
Huh? if you install anything via a software manager which is included with most user-friendly distros like Ubuntu, popos, mint or zorin, it comes with a .desktop file which makes it discoverable by using the means of the desktop environment - usually something like the start menu. And that's not something new. That has been the case for years now.
Companies don't really use Debian or Rocky in widescale production because they have no support.
Now red hat or ubuntu is a different matter.
Honestly though this does point out that this is a pattern of behavior on crowdstrikes part. This should have been the canary in the coalmine.
We actually use rocky and I think Debian at work for servers. We are currently migrating away from EOL centos .
We use Alma, which is basically Rocky. Before that, CentOS. Lots of people don't need or want the expensive support contracts.
OSS support though donations and commits is the way to go unless you get value out of those contracts (we would not).
I don’t know about that. In the HPC space we use a lot of EL distros. Mainly Centos & now Rocky. Most of the nodes run the os in ram too. Though almost all those kind of systems have no internet connection and don’t use things like crowdstrike. I’ve worked for a few places where the only part of the company that used windows was the office staff eg accounting, hr, etc. everything else is/was using an EL distro or upstream of one eg Fedora. Those type of places usually don’t mess things like crowdstrike for a lot of different reasons eg the kind of data they’re processing and security requirements on that data.
In April, a CrowdStrike update caused all Debian Linux servers in a civic tech lab to crash simultaneously and refuse to boot.
And then, you boot their servers from a Linux Live USB, run TimeShift to restore the last system snapshot, refuse the latest patch from Cloudstrike and they all lived happily ever after.
And it's not much more difficult to fix on Windows, except for the scale of the problem.
Good luck doing that remotely. Which is the sole problem with this most recent CrowdStrike bug.
Anybody who doesn't already have ipmi serial console access set up needs to put that on their list of acceptance criteria for remediation of this incident.
And on Windows you booted in safe mode and removed one file. What's the point of your post?
boot their servers from a Linux live usb
If I ran a computer lab that wasn't already net booted, I'd use this as the motivating factor to put that in place. Net booting to a repair image, or just reinstalling the whole OS either from scratch or a known good disk image, is where anybody who manages a fleet of computers should be.
There was a point in time where I had a pxe boot server vm set up on my laptop that I used to reload servers in our little row of racks at 365 main, because it let me quickly swap out the boot iso, and was faster than usb sticks were at the time.
Linux admins know that you're worsening security when installing 3rd party stuff into kernel, so most of them tend to avoid it. And that's why no one noticed that Crowdstrike problem.
Microsoft already has a very bad reputation, so they will be blamed for every issue on their OS.
Vista suffered from bad 3rd party drivers, then people proceeded to just dunk on M$ due to their already bad name. Despite Edge is nowadays just a different flavor of Chromium, people are still making "haha IE slow" memes, even those that still claim Google being the "savior of the internet".
So in the end, they is an internal contradiction in capitalism. It just append to be collapse due to lack of ressources and dumb management
I think that an economy lead by financial interest, open market, and a hierarchy in the production is a good definition of capitalism.
And yes, definitely the way that people get food, housing, and not being exclude will define a lot of thing in society.
Capitalism is, in essence, the ability for people to exchange their goods freely. It isn't dependent on corporations or some weird hierarchy of managers and workers. Those are facts of living in this system, but it isn't a direct consequence of "capitalism." If everyone worked only for themselves and produced something to bring to the exchange, that would still be capitalism.
It just append to be collapse due to lack of ressources and dumb management
TIL reverting the direction of Siberian rivers and turning Kazakh steppe into agricultural land were capitalist projects.
This one is a contradiction of highly hierarchical and degenerate systems.
With capitalism the contradiction is old and well known - power bends rules. Bent rules cause degeneracy. Degeneracy causes degradation and collapse.
Got me interested enough to Google, maybe you should too
Research and planning work on the project started in the 1930s and was carried out on a large scale in the 1960s through the early 1980s. The controversial project was abandoned in 1986, primarily for environmental reasons, without much actual construction work ever done.
If you mean the rivers part, then yes.
If you mean the steppe part, then no, they've caused a few ecological catastrophes first before stopping.
So who do you think hacked the DNC and got their emails, then? Is it the same people who hacked the RNC but didn't leak the emails? What makes you more qualified than CrowdStrike on this?
U.S. intelligence officials cannot make definitive conclusions about the hacking of the Democratic National Committee computer servers because they did not analyze those servers themselves. Instead, they relied on the forensics of CrowdStrike, a private contractor for the DNC that was not a neutral party, much as “Russian dossier” compiler Christopher Steele, also a DNC contractor, was not a neutral party. This puts two Democrat-hired contractors squarely behind underlying allegations in the affair – a key circumstance that Mueller ignores.
https://www.theguardian.com/us-news/2020/aug/18/donald-trump-us-senate-report-russia-campaign
A report by the Senate intelligence committee... runs to nearly 1,000 pages and goes further than last year’s investigation into Russian election interference by special prosecutor Robert Mueller... identifies Konstantin Kilimnik as a Russian intelligence officer employed by the GRU, the military intelligence agency behind the 2018 poisoning of the Russian double agent Sergei Skripal. It cites evidence – some of it redacted – linking Kilimnik to the GRU’s hacking and dumping of Democratic party emails.
I recently learned that this is the same company that gave us the bs Russia Gate.
WTF you mean the US Senate?
https://www.theguardian.com/us-news/2020/aug/18/donald-trump-us-senate-report-russia-campaign