Global BSOD

The fact that a software update not from MS can cripple the entire OS from even booting is just really crappy design. I don't care whether this is a server OS or a consumer one it really shouldn't happen. MS needs to make the OS far more resilient to this in the future.

Drivers for every operating system can do this. MS gives greater liberty to services like CrowdStrike - and that should be looked at - but every OS has vulnerabilities.
 
One of the things I liked about RISC OS - could always boot unstrapped from the ROM if the software environment got stuffed up in some way. Albeit that is more complicated when you have a larger range of possible base hardware configurations and require 3rd party drivers, etc.

What I do think this demonstrates is how pants on head stupid MS's (and other parties) approach to updates and automatic updates, especially forced, is, even in corporate environments which roll out stable tested images on more critical systems.
 
Last edited:
Question is why? Why this specific update? Or is it "common" practise to push updates bypassing QA testing? Again, why? Either way, a colossal boo boo.
I've spoken to ex-colleagues today and it has, as expected, been an utter nightmare alongside bitlocker'd end-points. Never been so glad to be retired!

e: gibberish on first line.

its only common practice for companies running waffer thin IT departments. I work in Pharma and my compnay was totally unaffected because we don't just blindly accept all security updaters. All updates get rinsed in snadbox first using clones of our production encironments ebfore anyone even thinks about signing off approvals to push to production. the reaosn its such a big problem is becuase of a combination of greed and laziness. A lot of big corporations don't bother even sandboxing windows updates and jsut let them get auto pushed to live environments for the sole reason that its cheap and easy. Why pay for someone to look at testing something that comes from a big multinational conglomorate ?????
 
its only common practice for companies running waffer thin IT departments. I work in Pharma and my compnay was totally unaffected because we don't just blindly accept all security updaters. All updates get rinsed in snadbox first using clones of our production encironments ebfore anyone even thinks about signing off approvals to push to production. the reaosn its such a big problem is becuase of a combination of greed and laziness. A lot of big corporations don't bother even sandboxing windows updates and jsut let them get auto pushed to live environments for the sole reason that its cheap and easy. Why pay for someone to look at testing something that comes from a big multinational conglomorate ?????

Im intrigued, your essentially saying your company tests every av definition update? bearing in mind these get released at least once a day with more on occasion? not many company's have that sort of ability (they may well invest in more now after this) and your very much in the minority if you are testing such updates.

One thing for sure is a lot of company's BCP's are going to be getting a refresh! we were not even hit but we will be adding a recovery plan for such an incident going forward. it might just involve shipping a usb stick with a windows installation on and letting autopilot do the rest but this incident has opened the eyes of a lot of organisations no doubt.
 
Last edited:
Correction.
It doesn't say which specific hospital had the issue just "A UK hospital" however I would bet money it's in England :P

Also, one hospital is a lot more than I would have expected to have issues with their Varian kit over this considering Varian's upgrade cycle speed (they're currently still working on transitioning their kit from Windows 7 to 10, no joke).
 
Drivers for every operating system can do this. MS gives greater liberty to services like CrowdStrike - and that should be looked at - but every OS has vulnerabilities.

While this is true and indeed Mac has had some issues with previous OS releases and stuff like Adobe software has caused some crashes before it doesn't have the BSOD thing Windows has.

If a MacBook crashes and someone is trying to access a GP appointment system or whatever then just restart it. More rarely they have kernel panic but again... reboot. The BSOD thing here required people to know to reboot in safe mode, delete a system file etc.. and often they don't have access to that or need to faff around and wait for IT end result GP receptionists, train people etc.. all basically locked out of their windows machines until IT sorts it.
 
It doesn't say which specific hospital had the issue just "A UK hospital" however I would bet money it's in England :p

Also, one hospital is a lot more than I would have expected to have issues with their Varian kit over this considering Varian's upgrade cycle speed (they're currently still working on transitioning their kit from Windows 7 to 10, no joke).

Specifically
Beatson West of Scotland Cancer Centre (part of the NHS)
 
its only common practice for companies running waffer thin IT departments. I work in Pharma and my compnay was totally unaffected because we don't just blindly accept all security updaters. All updates get rinsed in snadbox first using clones of our production encironments ebfore anyone even thinks about signing off approvals to push to production. the reaosn its such a big problem is becuase of a combination of greed and laziness. A lot of big corporations don't bother even sandboxing windows updates and jsut let them get auto pushed to live environments for the sole reason that its cheap and easy. Why pay for someone to look at testing something that comes from a big multinational conglomorate ?????



Including AV updates? Or just windows updates which are easier to test outside of production? I'm not familiar with Crowdstrike but does it have a install time offset? If it does, it looks like no one has implemented it.

Was curious so found this. https://www.youtube.com/watch?v=3cNh-au2g2g I expect many companies are going to change to n-1 or 2 after this!
 
Last edited:
Does anyone know how I can force BitLocker to prompt for a drive key for unlocking? Currently the machines are going straight to the CMD window without prompting to unlock the C: drive, and the drive is then invisible within the CMD window.

TIA
 
Does anyone know how I can force BitLocker to prompt for a drive key for unlocking? Currently the machines are going straight to the CMD window without prompting to unlock the C: drive, and the drive is then invisible within the CMD window.

TIA

Is the drive visible in diskpart? (dunno if you can then use mountvol).
 
Last edited:
its only common practice for companies running waffer thin IT departments. I work in Pharma and my compnay was totally unaffected because we don't just blindly accept all security updaters. All updates get rinsed in snadbox first using clones of our production encironments ebfore anyone even thinks about signing off approvals to push to production. the reaosn its such a big problem is becuase of a combination of greed and laziness. A lot of big corporations don't bother even sandboxing windows updates and jsut let them get auto pushed to live environments for the sole reason that its cheap and easy. Why pay for someone to look at testing something that comes from a big multinational conglomorate ?????
Pretty wild. I also worked in Pharma and the above comments certainly do not ring true for the company I contracted for, at least not for AV-type updates. I would think an incoming BCP review of all Windows based systems and offsetting / staggering future deployments of said updates. I worked in the cloud space and virtualisation, so fortunately would not be on the hook for significant recoveries if I was still in post but goodness me; the amount of noise this would have produced, endless hours of being in SWAT and side-bar calls. Ugh.

Good luck out there!
 
While this is true and indeed Mac has had some issues with previous OS releases and stuff like Adobe software has caused some crashes before it doesn't have the BSOD thing Windows has.

The sad Mac screen and the BSOD have the same function and purpose. Macs are mostly Unix underneath so they're better architected than Windows (although MS has got better) and they don't need to support the same diversity of hardware so you see the sad Mac less.

If a MacBook crashes and someone is trying to access a GP appointment system or whatever then just restart it. More rarely they have kernel panic but again... reboot. The BSOD thing here required people to know to reboot in safe mode, delete a system file etc..

That isn't down to a BSOD thing - you can normally boot straight back up after BSOD too, in fact it normally happens automatically - but this nuked the boot up process itself.
 
Back
Top Bottom