7 ms·

The article is pay walled. Seems like this would be the fault of the airlines though. There is a reason to be distributed between different geographic areas.

tanelpoder 2y ago
But if the Azure outage is due to Windows machines crashing because of the currently ongoing CrowdStrike crash/reboot loop issue, then such servers might end up being down in all regions. Looks like there might be some advanced lessons to be learned about blast radius here...
- switch007 2y ago
  Azure VM host machines are running CrowdStrike...??
  - bugbuddy 2y ago
    Maybe because Windows Defender Advanced Threat Protection is an enormous resource hog that scans every byte of memory and storage accessed by the Hypervisor and performs a quadratic time computation on the data? I am just guessing because my “fastest” Windows laptop CPU money could buy feels like a hot smelting furnace and a sloth at the same time when I use VMWare Workstation. What the &$@* is it scanning the VMWare guests for?
    - pjc50 2y ago
      Oh, Crowdstrike is also a massive resource hog.
  - tanelpoder 2y ago
    Probably not, but maybe they rely on some Active Directory server that is running it?
    - jaggederest 2y ago
      VM storage is probably on Windows Server, plus AD. I'd bet out of band management is all in the impact zone too. Might be back to someone pushing physical switches and hooking up a KVM.
    - mnw21cam 2y ago
      And this explains why the (Linux) HPC that I'm using is having so many troubles today, and keeps complaining that I don't exist.
  - pjc50 2y ago
    IF they're running windows, they'll be running it.
  - mgrund 2y ago
    More likely crash looping of so many VMs overloading some system with insufficient back pressure, possibly combined with unfortunate cluster management scheduler behavior at this scale of crash looping (e.g. too eager to retry scheduling instances, maybe even on new hosts which causes more infrastructure load).
- politelemon 2y ago
  It's more coincidental than likely, since these are managed services that were down, while the crowdstrike issue is closer to company deployments. But never say impossible, it's best to wait and see what the actual problem is rather than throwing shade so early.
  - politelemon 2y ago
    Update. Throw shade. They've confirmed it. I was wrong.

tanelpoder 2y ago

But if the Azure outage is due to Windows machines crashing because of the currently ongoing CrowdStrike crash/reboot loop issue, then such servers might end up being down in all regions. Looks like there might be some advanced lessons to be learned about blast radius here...

switch007 2y ago
Azure VM host machines are running CrowdStrike...??
- bugbuddy 2y ago
  Maybe because Windows Defender Advanced Threat Protection is an enormous resource hog that scans every byte of memory and storage accessed by the Hypervisor and performs a quadratic time computation on the data? I am just guessing because my “fastest” Windows laptop CPU money could buy feels like a hot smelting furnace and a sloth at the same time when I use VMWare Workstation. What the &$@* is it scanning the VMWare guests for?
  - pjc50 2y ago
    Oh, Crowdstrike is also a massive resource hog.
- tanelpoder 2y ago
  Probably not, but maybe they rely on some Active Directory server that is running it?
  - jaggederest 2y ago
    VM storage is probably on Windows Server, plus AD. I'd bet out of band management is all in the impact zone too. Might be back to someone pushing physical switches and hooking up a KVM.
  - mnw21cam 2y ago
    And this explains why the (Linux) HPC that I'm using is having so many troubles today, and keeps complaining that I don't exist.
- pjc50 2y ago
  IF they're running windows, they'll be running it.
- mgrund 2y ago
  More likely crash looping of so many VMs overloading some system with insufficient back pressure, possibly combined with unfortunate cluster management scheduler behavior at this scale of crash looping (e.g. too eager to retry scheduling instances, maybe even on new hosts which causes more infrastructure load).
politelemon 2y ago
It's more coincidental than likely, since these are managed services that were down, while the crowdstrike issue is closer to company deployments. But never say impossible, it's best to wait and see what the actual problem is rather than throwing shade so early.
- politelemon 2y ago
  Update. Throw shade. They've confirmed it. I was wrong.