This document provides guidance on collecting GPU instance logs to assist in analyzing and resolving GPU-related issues. The following instructions outline how to effectively collect these logs.
You can analyze the collected logs yourself or submit them to Tencent Cloud engineers for troubleshooting.
Retrieving Sub-instance dmesg and Serial Port Logs
Execute the command on the user instance:
Collecting NVIDIA GPU Logs
On a system with GPU drivers installed, execute the following command as the root user in any directory:
After the command is executed, a compressed log file named nvidia-bug-report.log.gz will be generated in the current directory.