Host System Diagnostics

Beyond routine monitoring and log review, the BMC provides advanced diagnostic tools designed to capture deep system-level data during and after a failure. These features are typically used when working with support engineers to diagnose complex hardware or firmware issues.

This chapter introduces the autonomous tools available for collecting debug data, generating crash dumps, and enabling large-scale diagnostics.

Configuring autonomous data collection (ADDC)

Autonomous Debug Data Collection (ADDC) allows you to configure how the system collects detailed debug information during an error recovery event, such as a warm or cold reset.

To configure ADDC:

1. In the sidebar menu, navigate to Host System Diagnostics > ADDC.

2. Configure the available options:

   • Error Recovery Mode: Select the system's response after an error: None, Warm Reset, or Cold Reset.

   • Data Harvesting Options:

     • Harvest uCode: Collects microcode debug data from the processor.

     • Harvest PPIN: Collects the Platform Processor Identification Number, a unique identifier for the CPU.

   • Polling Configuration: Proactively check for hardware errors at a defined interval.

     • MCA Polling: For Machine Check Architecture events.

     • DramCecc Polling: For DRAM Correctable ECC memory errors.

     • PCIeAer Polling: For PCIe Advanced Error Reporting.

   • Retry: Defines how many times the system should attempt to collect data if the first attempt fails.

3. Click Save.

[Image, EXISTING, Source: 10.1: Screenshot of the ADDC configuration page.]

Configuring autonomous crash dumps (ACD)

Autonomous Crash Dump (ACD) automatically generates and stores a crash dump file when the system detects a critical failure. This provides a snapshot of the system's state at the time of the crash, which is invaluable for root cause analysis.

To manage ACD:

1. Navigate to Host System Diagnostics > ACD.

2. Use the ACD Server toggle to enable or disable the automatic dump service.

3. To manually trigger a dump, click the Generate button.

4. Generated logs will appear in the table, where you can download them.

[Image, EXISTING, Source: 10.2: Screenshot of the ACD page with the Generate button.]