Google Compute Engine
Virtual machines running in Google's data centers
Exam Domain
Deploying and implementing cloud solutions represents approximately 23% of the exam. Compute Engine is heavily tested.
Machine Types
| Type | Use Case | Characteristics |
|---|
| E2 | General purpose, cost-optimized | Shared-core available, best price-performance |
| N2/N2D | General purpose, balanced | Intel/AMD processors, sustained use discounts |
| C2/C2D | Compute-intensive | Highest per-core performance |
| M2/M3 | Memory-optimized | Up to 12TB RAM, SAP HANA certified |
| A2/G2 | GPU workloads | NVIDIA GPUs for ML/rendering |
| Custom | Flexible sizing | Specify exact vCPUs and memory |
Preemptible vs Spot VMs
- Preemptible VMs: Up to 80% discount, max 24-hour runtime, 30-second shutdown notice
- Spot VMs: Same pricing as preemptible, no 24-hour limit, dynamic pricing
- Use Cases: Batch processing, fault-tolerant workloads, distributed computing
Instance Groups
| Type | Description | Use Case |
|---|
| Managed (MIG) | Identical VMs from template | Autoscaling, autohealing, rolling updates |
| Unmanaged | Different VM configurations | Legacy, heterogeneous workloads |
Exam Focus Areas
- Use Managed Instance Groups for stateless, scalable workloads
- Sustained use discounts apply automatically (up to 30% for month-long use)
- Committed use discounts require 1 or 3-year commitment for up to 57% savings
- Custom machine types allow exact CPU/memory specification
- Use startup scripts or custom images for consistent configuration
