You're correct.
— christian_taillon (@christian_tail) July 19, 2024
Full of zeros at least. pic.twitter.com/PJcCsUb9Vc
the big IT outage was caused by the r-slurs at crowdstrike pushing an update that contained a driver full of null bytes
https://x.com/christian_tail/status/1814299095261147448
- 22
- 60
Jump in the discussion.
No email address required.
I'm not an ITchad, but I'd like to learn more.
What does that mean in real terms?
Jump in the discussion.
No email address required.
tldr a company released an update that blanked out an incredibly important piece of code and now computers wont boot
less tldr, crowdstrike is a company that protects businesses from ransomware and viruses. they do this with a program called falcon, which uses a piece of code thats runs the moment you boot your computer. whats happening is this:
computer boots -> tries to read this piece of code -> freaks the frick out because the code is blank and it doesn't know what to do -> reboots
Jump in the discussion.
No email address required.
Thank you, that makes it clear. I wonder how something like that could possibly be allowed to happen.
Jump in the discussion.
No email address required.
I'm sure the devs at Crowdstrike are wondering that too
Jump in the discussion.
No email address required.
More options
Context
Complacency because previous updates worked
Process focused on quickly producing changes
Completely ignoring test automation because that's gay and that's what we hired Ranjeet (incompetent, thirdie) for
Jump in the discussion.
No email address required.
Used to be in test automation. All the shit US devs hated us because their shit code wouldn't pass. "Uhhhh but how do we know it wasn't your tests failing?" Because it's a five-line stack trace, you fricking r-slur.
As I was on my way out, they were replacing us with a bunch of Hyderabad devs. I hesitate to know the state of whatever the frick they put out now.
Jump in the discussion.
No email address required.
Devs should be checking in tested code. It's in their selfish interest and also it makes for less error-prone working conditions. Unfortunately that's an ideal and a lot of devs I come across don't think that way, but then this gay bullshit happens and it's kinda stressful and then people wish something had been done
Jump in the discussion.
No email address required.
It's a tough problem to solve because there is a twofold problem of devs fricking up when times are good and management fricking up when times are tough.
My favorite, of course, is management reacting to that crossroad with "where is the quality?" while playing annoying office politics that deter anyone from speaking up.
I bully other devs often, but bullying management is a dangerous game.
Jump in the discussion.
No email address required.
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context
More options
Context
More options
Context
He missed out the best part:
The computers downloaded the bad code
The computers can't boot any more
The computers can't download the fixed code
Fleets set up to re-image over netboot or something like that could be fixed without too much trouble, but most machines hit with this are going to need someone to sit down in front of them and fix them manually
Could take days just to travel to some of them
Jump in the discussion.
No email address required.
More options
Context
More options
Context
God bless my IT team. Got like 7 emails in the middle of the night about it and the last one was "all good now, if you do have issues just contact us". We were down for a bit and my guys on night shift went through heck for a few hours but not too bad honestly
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context