Send a diagnostic interrupt to debug unresponsive compute instances

  • Services: Compute
  • Release Date: Sept. 10, 2020
  • API Versions Affected: 20160918

You can debug an unresponsive or unreachable compute virtual machine instance by sending a diagnostic interrupt. On a configured instance, the diagnostic interrupt causes the OS to crash and generate a copy of the system memory, called a crash dump. You can analyze the crash dump to identify and debug the issue. For more information, including how to configure the OS to generate a crash dump, see Sending a Diagnostic Interrupt.