System information extraction from console output

[anatoly-savchenkov]

It would be useful if ESP-KVM could automatically collect system information from the target’s console output, both during boot and while the system is running.

A good amount of useful information is already visible on the KVM console without installing any agent on the target. During POST, bootloader and OS startup, ESP-KVM could build a basic inventory and health history from what appears on screen, for example:

• motherboard/system model and firmware/BIOS version
• CPU model and detected core/socket count
• installed memory and memory-related warnings
• storage controllers and detected disks/NVMe devices
• NICs, MAC addresses and PXE information
• boot device and bootloader/OS being started
• Secure Boot / TPM-related status when displayed
• OS name, kernel/version and hostname
• assigned IP addresses when shown during boot or on a console
• degraded RAID, disk, filesystem or mount errors
• hardware/firmware warnings and POST errors
• kernel panic, boot failures, emergency/rescue mode and other critical messages

The information could remain available in ESP-KVM as a simple target summary and boot history, making it possible to see what hardware and OS a machine has, what changed between boots, and why the last boot may have failed — even when the target itself is unreachable over the network.

This would be particularly valuable because it requires no agent, API or credentials on the managed system: it uses information that ESP-KVM can already observe through the console.