CrowdStrike effectively bricked windows, Mac and Linux today.
Windows machines won’t boot, and Mac and Linux work is abandoned because all their users are on twitter making memes.
Incredible work.
CrowdStrike effectively bricked windows, Mac and Linux today.
Windows machines won’t boot, and Mac and Linux work is abandoned because all their users are on twitter making memes.
Incredible work.
Gonna try my best here:
Crowdstrike is an anti-virus program that everyone in the corporate world uses for their windows machines. They released a update that made the program fail badly enough that windows crashes. When it crashes like this, it tries to restart in case it fixes the issue, but here it doesn’t, and computers get stuck in a loop of restarting.
Because anti-virus programs are there to prevent bad things from happening, you can’t just automatically disable the program when it crashes. This means a lot of computers cannot start properly, which means you also cannot tell the computers to fix the problem remotely like you usually would.
The end result is a bunch of low level techs are spending their weekends manually going to each computer individually, and swapping out the bad update file so the computer can boot. It’s a massive failure on crowdstrikes part, and a good reason you shouldn’t outsource all your IT like people have been doing.
It’s also a strong indicator that companies are not doing enough to protect their own infrastructure. Production servers shouldn’t have third party software that auto-updates without going through a test environment. It’s one thing to push emergency updates if there is a timely concern or vulnerability, but routine maintenance should go through testing before being promoted to prod.
It’s because this got pushed as a virus definition update and not a client update bypassing even customer staging rules that should prevent issues like this. Makes it a little more understandable because you’d want to be protected against current threats. But, yeah should still hit testing first if possible.
If a company disguises a software update as a virus definition update, that be a huge scandal and no serious company should ever work with them again…are you sure that’s what happened?
It wasn’t a virus definitions update. It was a driver update. The driver is used to identify and block threats incoming from wifi and wired internet.
The “Outage” section of the Wikipedia article goes into more detail: https://en.wikipedia.org/wiki/2024_CrowdStrike_incident#Outage
100% agree. I haven’t been on the backend of managing crowdstrike so I don’t know if this is a option, but running a wsuz server and manually weeding out bad updates was such an improvement over rawdogging windows updates.