- We updated definitions file (but we call it something cooler) and it blew up the kernel.
- We update this file multiple times a day and we're not going to stop doing it.
- The file you have to delete isn't a driver! It totally isn't read by a driver that completely fucking broke the kernel!
- The file is just evaluating every fucking pipe in the system no biggie.
- We haven't decided who to throw under the bus yet even though the only way we could have done a fuckup this big was to have multiple policies and procedures fail massively or not exist in the first place.
- Good luck!
Any company that does not dump crowdstrike after this deserves to have another outage
Also like, this is the warning for every IT department out there. You have no excuses to not take action from Monday forward.
If you let someone else control how and when your systems update, and they're not responsible for personally going out to unfuck your systems if they push something bad, you're asking for major outages. Doesn't matter if it's crowdstrike, microsoft, or fucking canonical. They will push things out untested. They will use you as a test environment. They will not help when they break your stuff.
Unless you'd rather do the "we made sure our contract says if you break something we're going to be paid well, and we have an in house legal department to make sure of that" thing, I guess.