Alibaba ACA Cloud Business study plan

Map each of the four exam domains to its weight and rank them by study priority

List the core Alibaba Cloud compute families (ECS instance types, images, disks) and describe when each is chosen

Configure a VPC with subnets, route tables, and a security group in a sandbox account

Explain the shared responsibility model and the account/identity hierarchy (RAM users, roles, policies)

Compare OSS storage classes and lifecycle rules, then apply a lifecycle policy to a test bucket

Differentiate SLB, ALB, and CLB by use case and sketch a load-balancing topology

Describe RDS vs PolarDB vs AnalyticDB positioning and pick one for three sample workloads

Summarize CloudMonitor and Log Service capabilities for basic observability

Diagram a multi-tier web architecture with VPC segmentation and load balancing

Justify region and zone placement decisions for availability and latency

Design a backup and disaster-recovery approach with stated RPO/RTO targets

Apply least-privilege RAM policies to a sample multi-role architecture

Identify single points of failure in a given design and propose redundancy fixes

Size compute and storage for a stated workload and defend the estimate

Classify migration approaches (rehost, replatform, refactor) against three sample applications

Produce a cost-aware design tradeoff table for two candidate architectures

Build a phased migration plan with assessment, pilot, cutover, and rollback stages

Draft a cutover checklist including data sync validation and DNS switch steps

Map common on-premises components to Alibaba Cloud equivalents

Define success metrics and a post-migration validation procedure

Describe the PAI platform components and where each fits in an ML lifecycle

Explain model training vs inference infrastructure needs and select instance types for each

Outline a generative AI solution pattern using managed model APIs with a retrieval layer

Identify data governance and security controls required for AI workloads

Compare prompt engineering, RAG, and fine-tuning by cost, latency, and accuracy tradeoffs

Design a RAG pipeline with ingestion, chunking, embedding, and retrieval stages

Specify evaluation criteria and a test set for a generative AI feature

List cost-control and rate-limiting tactics for inference workloads

Re-derive the domain weights and confirm weak areas from practice results

Solve timed practice questions across all four domains and log error patterns

Rewrite the weakest three concept summaries from memory

Review all prior deliverables and close remaining gaps