Since early this morning, six of our servers have suddenly become completely unresponsive and required a hard reboot. Most of the affected machines are hosted in AWS, but at least one on-premises server has had the same problem. Windows event logs do not show anything obvious immediately before the freezes. We have not installed recent Windows patches because updates are currently on hold due to known bugs. Is anyone else experiencing similar lockups, or have ideas for what infrastructure or software issue might be causing them?
2 Answers
This could be related to a recent update or hotfix, even if the normal Windows patches were deferred. A newer Hyper-V host gave me similar trouble this morning, so I would compare affected machines and hosts, then check whether any security or endpoint updates were deployed independently of Windows Update.
Since most of the affected systems are in AWS, check the account-specific health dashboard and the availability zones involved. Problems at the rack, row, or host level can sometimes make a group of instances appear frozen. A stop-and-start cycle may place an instance on a different underlying host, which can help confirm whether the issue is infrastructure-related.
That is worth checking, although the issue has also affected a server in one of our on-premises data centers, so it may not be limited to AWS.
We have not pushed any recent Windows updates, though we are checking whether an endpoint security hotfix or similar change was installed separately.