Kernel Panic

A kernel panic is basically the Linux kernel saying, "I hit a critical error and can't safely continue."
It's like the Linux equivalent of the Windows Blue Screen of Death (BSOD).

Some conditions trigger an immediate panic, while others allow the system to recover or keep going.

Kernel Panic Settings (sysctl)

These sysctl settings control how the system reacts when specific failures occur.

sysctl -a | grep -i 'panic'

Setting	Value	What it means
kernel.panic	0	After a panic, the system will not automatically reboot.
kernel.panic_on_oops	0	The system will not panic if a kernel "oops" (non-fatal bug) occurs.
kernel.hardlockup_panic	1	If a CPU lockup is detected (e.g. no heartbeat), the system panics.
kernel.hung_task_panic	0	Long-running/stuck tasks won’t cause a panic.
vm.panic_on_oom	0	The system will not panic if it runs out of memory (OOM).
kernel.panic_on_warn	0	Kernel warnings (via WARN()) do not trigger panic.
kernel.panic_on_rcu_stall	0	RCU (Read-Copy-Update) stalls won't cause panic.
kernel.panic_on_unrecovered_nmi	0	Unrecoverable NMIs won't panic the system.
kernel.unknown_nmi_panic	0	Unknown NMIs also won’t cause a panic.
kernel.panic_print	0	Controls what info is printed on panic — 0 means minimal info.