Answer Key: AlloyDB

Back to Exercises

Exercise 1: Design Architecture

Question: Design an AlloyDB architecture for a high-traffic application. How many read replicas? What's the failover strategy?

Answer

Goal: Design high-availability AlloyDB architecture.

Architecture

Primary Instance: - Region: us-central1 - Machine Type: db-standard-8 (8 vCPU, 30GB RAM) - High Availability: Enabled

Read Replicas: - Replica 1: us-central1 (same zone as primary) - Replica 2: us-east1 (different region) - Replica 3: us-west1 (different region)

Failover Strategy: - Automatic failover (< 60 seconds) - Promote read replica to primary - Update application connection strings

Answer: Use 3 read replicas (1 same region, 2 different regions), enable automatic failover, design for < 60 second RTO.

Exercise 2: Handle Scaling

Question: Your application needs to scale reads. How do you scale with AlloyDB? What's the strategy?

Answer

Scaling strategies:

  1. Add Read Replicas: Add more read replicas for read scaling
  2. Connection Pooling: Use connection pooling for efficient connections
  3. Read-Write Splitting: Route reads to replicas, writes to primary
  4. Monitor Performance: Monitor replica lag and performance

Answer: Add read replicas for horizontal read scaling, use connection pooling, implement read-write splitting.

Exercise 3: Optimize Performance

Question: Your queries are slow. How do you optimize them? What AlloyDB features do you use?

Answer

Optimization strategies:

  1. Use Columnar Cache: Enable columnar cache for analytics
  2. Optimize Queries: Use appropriate indexes and query patterns
  3. Connection Pooling: Use connection pooling
  4. Monitor Performance: Monitor query performance and optimize

Answer: Enable columnar cache, optimize queries with indexes, use connection pooling, monitor and optimize continuously.