dmesg: Reading Kernel Ring Buffer for Hardware and Boot Diagnostics

Hardware errors, driver failures, and OOM kill events are not visible in application logs — they appear only in the kernel ring buffer.

Use dmesg with filters and human-readable timestamps to diagnose kernel-level events.

# Basic dmesg with human timestamps
dmesg -T

# Follow kernel messages in real time
dmesg -Tw

# Filter by log level (err, warn, crit, etc.)
dmesg -l err,crit

# Filter since last boot
dmesg -T | head -50

# Search for specific hardware errors
dmesg | grep -i 'error\|fail\|oom\|killed'

# OOM killer events (process killed for out of memory)
dmesg | grep -i 'Out of memory\|oom_kill'

# USB device events
dmesg | grep -i usb

# Disk errors (SCSI/ATA)
dmesg | grep -i 'ata\|scsi\|error\|reset'

# Check if kernel messages are also in journal
journalctl -k -p err

dmesg reads the kernel ring buffer — a circular in-memory log. It contains events that happen before syslog starts (boot) and kernel-level events that bypass userspace logging entirely.