First major cloud provider to offer per-second billing for IaaS compute offering Compute Engine Also offers per-second billing for: - Google Kubernetes Engine - Dataproc - App Engine Sustained-use discounts Virtual resources can be configured to tailor pricing for workloads Setting budgets - fixed - tied to a metric Alerts: - generally set at 50%, 90%, and 100% of budget Reports: - visual tool to monitor expenditure of project or services Quotas: - prevent over-consumption of resources due to: - errors - malicious attacks - rate quota - reset after amount of time - 3000 API calls per 100 seconds - allocation quota - govern number of resources for the project - default: less than 15 Virtual Private Cloud networks