GCP Associate Cloud Engineer Practice Question 96: Serial Console VM Boot Failure
Google Cloud Associate Cloud Engineer · Test 10 · question 6 of 10
A Compute Engine VM instance stops responding to SSH after a kernel update was applied inside the guest OS, and the instance never becomes reachable again after a reboot. You cannot connect over SSH because the network stack inside the guest never finishes initializing. You need to see the boot-time messages the guest OS is producing to diagnose why it is failing to start correctly.
What should you do?
Show answer and explanation
Answer:
The serial console captures the output written to the VM's serial port, which includes kernel and boot-time messages from the guest OS even when the network stack never comes up and SSH is unreachable. Reading the serial port output (or connecting interactively to the serial console) is the standard way to diagnose why a VM is failing to complete its boot process.
Option A is wrong because Cloud NAT provides outbound internet access for instances without external IPs; it has no effect on a guest OS that is failing during its own boot sequence. Option B is wrong because moving the disk to a new VM in another zone does not reveal any boot diagnostics and does not address whatever the kernel update broke; the new VM would likely fail to boot the same way. Option C is wrong because attaching an unrelated empty disk does not trigger useful diagnostics for the boot disk's own kernel or filesystem problem.
Reference: Interacting with the serial console