Examkin

Question

Difficulty: MediumManaging Compute Engine Resources

An organization processes financial batch jobs using a Compute Engine Managed Instance Group (MIG) equipped with an autoscaler. During sudden traffic drops, the autoscaler immediately terminates instances, causing several active, long-running batch jobs to fail prematurely. The cloud engineering team must ensure active jobs complete safely while still allowing the MIG to scale down gracefully when load decreases. Which TWO actions should the team implement to meet these requirements?

  1. Configure scale-in controls within the MIG autoscaling policy and specify a stabilization period.Answer
  2. Apply instance protection from scale-in to specific Compute Engine instances while they are executing long-running jobs.Answer
  3. C
    Update the instance template used by the MIG to deploy Spot VMs to automatically block scale-in terminations.
  4. D
    Assign the primitive Owner role to the default Compute Engine service account to prevent automated deletion operations.
  5. E
    Submit a regional Compute Engine vCPU quota increase request via the Google Cloud Console to halt autoscaling scale-in operations.

Answer

To protect active long-running batch jobs during autoscaling scale-in events, the team should configure scale-in controls with a stabilization period in the MIG autoscaling policy and apply instance protection from scale-in on VM instances actively processing jobs.
Configuring scale-in controls with a stabilization period establishes a trailing evaluation window that prevents sudden scale-in spikes. Additionally, enabling instance protection from scale-in on specific active instances explicitly guarantees that the MIG autoscaler will not terminate those instances while they complete long-running batch jobs.

Step-by-Step Solution

1
Analyze the autoscaling behavior during scale-in events.
Identified that rapid autoscaler scale-in immediately deletes VM instances carrying active workloads.
Standard autoscaling scales down as soon as metrics drop unless scale-in parameters or protection flags are configured.
2
Configure MIG-level scale-in controls.
Setting a stabilization period forces the autoscaler to evaluate peak load over a trailing time window before removing instances.
Stabilization windows prevent abrupt VM terminations during momentary metric fluctuations.
3
Apply instance protection on target VMs.
Specific instances running critical jobs are shielded from deletion until their tasks finish.
Instance protection explicitly overrides autoscaler selection for deletion on designated compute instances.

Key Concept

Managed Instance Group Scale-in Control and Instance Protection
Rate this question

Topics

Managing Compute Engine Resources

More questions from this topic