Google ACE Practice Question: Ensuring Successful Operation of a Cloud Solution

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Delete the instance and recreate it with the new machine type.

  Why it's wrong here

  Deleting the instance is unnecessarily destructive and risks losing its unique configuration, such as custom metadata, tags, IAM policies, and ephemeral external IP. While the boot disk could be retained and reattached to a new instance, this approach requires manually recreating all settings and is far more error-prone than simply changing the machine type. The correct procedure never requires deletion; a straightforward stop, retype, and start preserves the instance identity, disks, and networking configuration.

• ✗

  Run gcloud compute instances set-machine-type while the instance is running.

  Why it's wrong here

  The gcloud compute instances set-machine-type command will fail at the API level if the instance is in the RUNNING state, because the machine type (vCPU and memory allocation) can only be modified while the instance is TERMINATED. The command does not perform an implicit stop; attempting it returns a Required 'machineType' to be specified or instance must be terminated error. Therefore, this option is incorrect because it skips the mandatory stop step, making the command ineffective on a live instance.

• ✓

  Stop the instance, run gcloud compute instances set-machine-type, then start the instance.

  Why this is correct

  This is the only correct sequence: first stop the instance (gcloud compute instances stop), which transitions it to the TERMINATED state, then call gcloud compute instances set-machine-type with the desired type (predefined, custom, or E2), and finally start the instance again. The stop/start cycle is required because the hypervisor must release the old vCPU and memory resources before the new allocation can be applied. After the instance starts, it retains its existing disks, IP addresses, and metadata, so no configuration is lost.

• ✗

  Take a snapshot, create a new instance, and attach the disk.

  Why it's wrong here

  Taking a snapshot and creating a new instance with the disk attached is unnecessary because the machine type of a running Compute Engine instance can be changed directly via a `gcloud compute instances set-machine-type` command or the console, requiring only a stop and restart. This option is tempting because snapshot-and-recreate is the correct procedure for migrating an instance to a different region or zone, where the disk must be recreated from a snapshot.

Go deeper

Related to this question