Search...

Professional Cloud Architect Study Notes

Compute Engine

IaaS virtual machines in GCP

Machine Types

Family	Use Case	vCPU:Memory
General Purpose (E2, N2)	Balanced workloads	1:4 GB
Compute Optimized (C2, C3)	HPC, gaming	1:2 GB
Memory Optimized (M2, M3)	In-memory databases	1:14+ GB
Accelerator Optimized (A2)	ML training	GPUs attached

Availability Options

Zonal

Single zone deployment, lowest cost

Regional MIG

Managed Instance Group across zones

Multi-region

Load balancer across regions

Pricing Models

On-demand: Pay per second, no commitment
Committed Use: 1-3 year commitment, up to 57% discount
Spot VMs: Up to 91% discount, can be preempted
Sustained Use: Automatic discounts for consistent usage

Exam Focus Areas

Use custom machine types for exact resource needs
Preemptible/Spot VMs for fault-tolerant batch jobs
Sole-tenant nodes for licensing or compliance requirements
Live migration keeps VMs running during maintenance

GKE & Cloud Run

Container orchestration and serverless containers

Google Kubernetes Engine (GKE)

Mode	Control Plane	Best For
Autopilot	Google-managed	Hands-off, pay per pod
Standard	Google-managed	Full node control
Anthos	Multi-cloud	Hybrid/multi-cloud Kubernetes

Cloud Run

Serverless

Scale to zero, pay per request

Revisions

Traffic splitting, canary deployments

Concurrency

Multiple requests per instance

Triggers

HTTP, Pub/Sub, Cloud Scheduler

App Engine

Standard: Language-specific runtimes, scale to zero
Flexible: Docker containers, minimum 1 instance
Automatic scaling based on traffic
Integrated with Cloud services

Exam Focus Areas

GKE Autopilot for simplified Kubernetes management
Cloud Run for stateless HTTP containers
Cloud Functions for event-driven single functions
App Engine for simple web apps without container knowledge

Read GCP-PCA notes without distractions

Open Foci to run a focused study block while you review domains, tables, and exam tips.

Focus while reading

Professional Cloud Architect Study Notes

Compute Engine

IaaS virtual machines in GCP

Machine Types

Family	Use Case	vCPU:Memory
General Purpose (E2, N2)	Balanced workloads	1:4 GB
Compute Optimized (C2, C3)	HPC, gaming	1:2 GB
Memory Optimized (M2, M3)	In-memory databases	1:14+ GB
Accelerator Optimized (A2)	ML training	GPUs attached

Availability Options

Zonal

Single zone deployment, lowest cost

Regional MIG

Managed Instance Group across zones

Multi-region

Load balancer across regions

Pricing Models

On-demand: Pay per second, no commitment
Committed Use: 1-3 year commitment, up to 57% discount
Spot VMs: Up to 91% discount, can be preempted
Sustained Use: Automatic discounts for consistent usage

Exam Focus Areas

Use custom machine types for exact resource needs
Preemptible/Spot VMs for fault-tolerant batch jobs
Sole-tenant nodes for licensing or compliance requirements
Live migration keeps VMs running during maintenance

GKE & Cloud Run

Container orchestration and serverless containers

Google Kubernetes Engine (GKE)

Mode	Control Plane	Best For
Autopilot	Google-managed	Hands-off, pay per pod
Standard	Google-managed	Full node control
Anthos	Multi-cloud	Hybrid/multi-cloud Kubernetes

Cloud Run

Serverless

Scale to zero, pay per request

Revisions

Traffic splitting, canary deployments

Concurrency

Multiple requests per instance

Triggers

HTTP, Pub/Sub, Cloud Scheduler

App Engine

Standard: Language-specific runtimes, scale to zero
Flexible: Docker containers, minimum 1 instance
Automatic scaling based on traffic
Integrated with Cloud services

Exam Focus Areas

GKE Autopilot for simplified Kubernetes management
Cloud Run for stateless HTTP containers
Cloud Functions for event-driven single functions
App Engine for simple web apps without container knowledge

Read GCP-PCA notes without distractions

Open Foci to run a focused study block while you review domains, tables, and exam tips.

Focus while reading