System Logs and Events

While real-time sensor data is great for spotting current issues, system logs provide a historical record of everything that has happened on the server. When you're investigating a past failure or an intermittent problem, the logs are the first place you should look.

This chapter teaches you how to read the different types of logs the BMC records, from high-level system events to low-level boot codes.

Viewing and filtering event logs

The Event Log is your primary source for understanding the server's health history. It captures hardware-related alerts, status changes, and errors, each with a timestamp and a severity level.

To view the event logs:

1. In the sidebar menu, navigate to Logs > Event logs.

   
2. You will see a table of all recorded events.

[Image, EXISTING, Source: 4.2: Screenshot of the Event Logs page, showing a list of critical, warning, and OK events.]

Each log entry includes:

• Severity: The importance of the event (e.g., Critical, Warning, OK).

• Date: The UTC timestamp when the event occurred.

• Description: A brief explanation of the event.

Pro Tip: How to approach the event log

When troubleshooting, filter the log to show only Critical and Warning events first. This helps you cut through the noise of normal operational messages (OK) and focus on the actual problems. You can also filter by date to narrow down events that occurred around the time of a reported failure.

To export the logs for sharing or offline analysis, click the Export all button. You can typically export them in formats like JSON, which is useful for sending to support teams.

Field Reference

Field	Description
ID	A unique identifier for each log entry.
Severity	The importance of the event (e.g., Critical, Warning, OK).
Date	The UTC timestamp when the event occurred.
Description	A brief explanation of the event that was triggered.
Status	Allows you to manually toggle the status to "Resolved" once the issue has been addressed.

Analyzing POST code logs for boot issues

If a server fails to boot and you can't even get to the OS, the POST (Power-On Self-Test) code log is an invaluable tool. It records the sequence of hardware initialization codes generated by the BIOS during startup.

Navigate to Logs > POST code logs to view these records.

[Image, EXISTING, Source: 4.3: Screenshot of the POST code logs page, showing a list of hexadecimal codes with timestamps.]

If the server hangs during boot, the last POST code recorded in this log can often tell you exactly which hardware component or initialization step is causing the problem.

Field Reference

Field	Description
Created	The timestamp when the POST code entry was recorded.
Time Stamp offset	The time offset, in seconds, relative to the start of the boot process.
Boot count	The boot cycle number associated with this code.
POST code	The hexadecimal value representing a specific hardware initialization checkpoint.

Generating and downloading system dumps

For complex, low-level issues, a support engineer might ask you for a "system dump." This is a snapshot of the BMC's internal state and memory, which can be used for deep-level debugging.

To generate a system dump:

1. Navigate to Logs > Dumps.

2. Select the dump type (usually BMC dump).

3. Click Initiate dump.

[Image, EXISTING, Source: 4.1: Screenshot of the Dumps page, showing the "Initiate dump" button and a list of previously generated dumps.]

The process can take a few minutes. Once complete, the dump file will appear in the list, and you can download it to your local machine.

Field Reference

Field	Description
Date and time	The timestamp when the dump was completed.
Dump type	The type of dump collected (e.g., BMC Dump Entry).
ID	A unique identifier for the dump file.
Size	The file size of the generated dump.

Reviewing video logs of system events

Some platforms are configured to automatically record a short video clip of the KVM console output when a critical event occurs (like a system reset or a crash). This can help you see exactly what was on the screen at the moment of failure.

Navigate to Logs > Video Log to see a list of available recordings. You can download and play these video files (video.dat) to aid your diagnosis.

[Image, EXISTING, Source: 4.4: Screenshot of the Video Log page with a single video file listed.]

Capturing Blue Screen of Death (BSOD) images

If the host operating system crashes (for example, showing a "Blue Screen of Death" on Windows), the BMC can automatically capture a screenshot of the failure screen. This is extremely useful because it preserves the error codes and messages displayed on the screen, which might otherwise be lost after a reboot.

To manage a captured BSOD image:

1. Navigate to Settings > BSOD.

2. The page will display the last captured screen.

3. You have three options:

   • Trigger: Manually initiates a capture of the BSOD screen if the host is in a critical stop state.

   • Download: Saves the currently displayed BSOD image to your local machine for analysis.

   • Delete: Clears the stored image from the BMC.

[Image, EXISTING, Source: 7.1: Screenshot of the BSOD page showing a captured screen from a host OS.]

Best Practice: Don't reboot too quickly

After a server crashes, resist the urge to immediately reboot it. Log into the BMC first and check the BSOD page. Capturing that screenshot can be the key to understanding the root cause of the crash.