Comments by "" (@grokitall) on "Global tech outage: Microsoft VP explains what went wrong" video.

  1. because a lot of these machines are locked down, it would not help. you cannot stop buggy kernel drivers from crashing the kernel, but you can put things in place to automatically disable them next time you boot, meaning no boot loop, just power cycle the machine and it is back up and running, and able to check for another update.
    2
  2.  @richardsauer1667  you missed the point. this was a kernel level update, so they could not prevent it crashing. however they could have created something to stop it from boot looping after the similar incident caused by mccaffee, where the current cloudstrike ceo was working at the time.
    1
  3. any os can be shutdown by a buggy kernel level driver, the question is how well does it recover when you reboot. in this case the answer is that it did not recover, which is firmly Microsofts fault.
    1
  4. this time, they crashed debian and rocky linux a few months ago, but they recovered better.
    1
  5.  @aaaa1953  good question. i wish there was a good answer. we know enough about writing software that such a broken driver should not be able to ship, and enough about writing kernels that if it shipped, it should not be able to boot loop and keep the kernels down like it did, so both microsoft and cloudstrike have some serious questions to answer, potentially in court.
    1