All GCP services

GCP · Compute

Compute Engine

Scalable VMs on Google's infrastructure.

Official docs

Overview

Compute Engine provides configurable Linux/Windows VMs with custom machine types, GPUs, sustained-use discounts and live migration.

When to use it

  • Lift-and-shift of on-prem workloads
  • Test agents, build runners, load generators
  • Workloads needing custom kernels or GPUs

Setup

  1. Enable Compute Engine API in your GCP project.
  2. Console → Compute Engine → Create instance.
  3. Pick machine type (e2-micro for free tier, n2-standard for prod).
  4. Choose boot image, network tags and firewall rules.
  5. Add SSH keys (project- or instance-level) and create.

How to use

Create a VM with gcloud
gcloud compute instances create qa-runner \
  --machine-type=e2-medium \
  --image-family=debian-12 --image-project=debian-cloud \
  --tags=http-server
SSH in
gcloud compute ssh qa-runner --zone=us-central1-a
Delete
gcloud compute instances delete qa-runner --zone=us-central1-a

QA use cases

  • Dedicated Selenium/Playwright grid nodes in a managed instance group.
  • k6/JMeter load generators in an autoscaling MIG.
  • Scheduled nightly regression VMs spun up by Cloud Scheduler + Cloud Functions.