Q: An incident occurred where a server ran out of memory, but the `oom-killer` didn't trigger, causing a complete kernel hang (hard lockup). How do you configure the kernel to automatically panic and reboot if this happens again?

If the system becomes completely unresponsive in a hard lockup or OOM stall, manual intervention is slow. I would configure the kernel to panic and automatically reboot. This allows auto-scaling groups or load balancers to immediately replace or re-route traffic from the failed node.

• kernel.panic = 10 (Reboot 10 seconds after a panic)
• vm.panic_on_oom = 1 (Panic if the kernel hits an unresolvable OOM state, rather than trying to kill processes if the killer is disabled or failing)
• kernel.hung_task_panic = 1 (Panic if tasks are hung for too long, like the D state lockups).

Using sysctl: These settings are added to /etc/sysctl.conf to persist across reboots, forcing the system to "fail fast" and rely on infrastructure redundancy.