'Largest IT Outage in History' Sparks Global Chaos.
'Largest IT Outage in History' Sparks Global Chaos.
"This is basically what we were all worried about with Y2K, except it's actually happened this time."
You're viewing a single thread.
So was this Crowdstrike's fuck up and not Microsoft's?
Probably, but the issue is in the interface between Windows and the CrowdStrike software causing Windows to go into a crashing bootloop.
Closed source is great, I tell you. /s
It has nothing to do with closed source, this is entirely about a privileged application fucking around and not testing shit before pushing it globally. You can have the same issues with Linux too. My org has stated that our AV product is never to be installed on Linux machines because they hosed tons of machines years back doing something similar.
High privilege security applications are always going to carry this risk due to how deeply they hook into the OS to do their job.
That is true. An obvious failure is that the update that broke everything was pushed everywhere simultaneously.
That's what has me confused. I haven't even stepped into an office in 20 years so I have no modern experience there, but I would have thought that a company with such a massive user base would release things at different times based on region. Is it because with security based applications they don't want to risk someone having time to exploit a new vulnerability?
Is it because with security based applications they don't want to risk someone having time to exploit a new vulnerability?
Pretty much. Given how fast the malware scene evolves and implements day 1 exploits, and how quickly they need to react to day 0 exploits, there's kind of an unwritten assumption (and it might actually be advertised as a feature) that security software needs to react as fast as possible to malicious signatures. And given the state of malware and crypto shit, it's hard to argue that it isn't needed, considering how much damage you'll suffer if they get through your defenses.
That being said, this kind of a fuck up is damned near unacceptable, and these updates should have been put through multiple automated testing layers to catch something like this before this got to the end user devices. I could see the scale of it taking them out of business, but I also wouldn't be surprised if they managed to scrape by if they handle it correctly (though I don't see the path forward on this scale, but I'm not a c-suite for many reasons). Like I said above, we had an incident years back that hosed a bunch of our Linux boxes, but the vendor is still around (was a much smaller scale issue) and we even still use them because of how they worked with us to resolve the situation and prevent it from happening again.