Amazon has introduced SendDiagnosticInterrupt for AWS EC2. The feature is an API-driven diagnostic interrupt that triggers a kernel panic for Linux instances or a blue screen/stop error for Windows instances. The interrupt is received as a non-maskable interrupt by Intel and AMD processors.
This feature is especially helpful for administrators managing cloud instances, where no hardware interrupt button is available to force fail a system. With the diagnostic interrupt triggered, administrators, can crash reboot the OS, as well as configure the OS to perform diagnostic tasks, memory dumps, secondary kernel loads, trace calls and more.
Amazon has published guides for both Linux and Windows to help administrators learn how to configure an OS to generate a crash dump in the event of a kernel panic or stop error. Parameters available for this action include DryRun (to check whether required permissions for the action are available) and InstanceID (for ID of the instance). Check out the reference docs to learn more and see examples.