database-architect

Database Architect - Database Architecture Design Expert

Skills Overview

Database Architect is a professional database architecture design assistant focused on building scalable, high-performance data layer architectures from scratch. Whether you are choosing database technologies, designing data models, or planning database migrations, it provides comprehensive architectural advice and best practices.

Applicable Scenarios

1. Database Technology Selection and Architecture Design

When you need to choose the right database technology for a new project, Database Architect can help analyze options such as SQL, NoSQL, time-series databases, and graph databases, and provide selection recommendations based on business scenarios, access patterns, and scale targets. Whether it’s relational databases like PostgreSQL and MySQL, or NoSQL solutions like MongoDB and DynamoDB, it offers clear recommendations and trade-off analyses.

2. Data Modeling and Schema Design

When you need to design database table structures, document models, or data partitioning strategies, Database Architect can help you through the full process from conceptual modeling to physical design. This includes entity-relationship design, trade-offs between normalization and denormalization, index strategy planning, multi-tenant data isolation schemes, and more—ensuring the data model meets business needs while supporting future growth.

3. Database Migration and Architectural Evolution

When you need to migrate an existing system to a new database architecture or transform a monolithic database into a microservices-oriented design, Database Architect can provide zero-downtime migration strategies, sharding plans, cross-region replication designs, and more. It supports cross-database migrations from SQL to NoSQL, as well as phased migration plans for large tables to ensure data safety and business continuity.

Core Features

1. Full-Stack Database Technology Evaluation

Covers mainstream technologies including relational databases (PostgreSQL, MySQL, SQL Server, Oracle), NoSQL (MongoDB, DynamoDB, Cassandra), time-series databases (TimescaleDB, InfluxDB, ClickHouse), NewSQL (CockroachDB, TiDB), and graph databases (Neo4j). Provides analysis of performance characteristics, operational complexity, and cost. Based on theoretical frameworks such as the CAP theorem and consistency models, it helps you make informed decisions when selecting technologies.

2. Data Modeling and Index Optimization

Offers a complete methodology from conceptual modeling to physical design, including entity-relationship diagrams, normalization (1NF-5NF), dimensional modeling, partitioning strategies, etc. Designs index strategies tailored to query patterns and access frequency, covering index types such as B-tree, Hash, GIN, BRIN, as well as composite indexes, partial indexes, JSONB indexes, and other advanced techniques. Also plans caching architectures, including Redis/Memcached selection, caching strategies (cache-aside, write-through), and eviction/expiration mechanisms.

3. Scalability and High Availability Design

Provides vertical and horizontal scaling strategies based on business growth forecasts, read-write separation, connection pool configuration, shard key selection, and other architectural designs. Plans high-availability solutions such as primary-replica replication, multi-primary replication, and multi-region deployments, and guides consistency model choices (strong consistency/eventual consistency). Also includes disaster recovery designs like backup strategies, point-in-time recovery, RPO/RTO planning, and failover procedures to ensure the resilience and reliability of the database architecture.

Frequently Asked Questions

How do I choose the right database technology for my project?

Choosing a database requires considering the business scenario, data model, access patterns, and expected scale. First, analyze your data characteristics: structured data fits relational databases (PostgreSQL, MySQL); semi-structured or flexible schemas fit document databases (MongoDB); highly connected social networks fit graph databases (Neo4j); time-series data fits TimescaleDB or InfluxDB. Next, assess consistency needs: choose ACID-compliant databases for strong consistency requirements; NoSQL options may be acceptable when eventual consistency is tolerable. Also consider operational complexity, team expertise, and cloud service integration. Database Architect will provide detailed selection comparisons and recommendations based on your specific requirements.

Should I normalize or denormalize in database design?

This is a trade-off decision. Normalization brings benefits like higher data consistency, efficient updates, and optimized storage, but may lead to complex JOIN queries. Denormalization can improve read performance and reduce query complexity, but increases data redundancy and maintenance costs. Recommended practice: prioritize normalization for OLTP (transactional) scenarios; use dimensional modeling (star/snowflake schemas) for OLAP (analytical) scenarios; selectively denormalize frequently accessed data in read-heavy, write-light scenarios; keep core tables that experience frequent writes normalized to maintain data consistency. Database Architect will design a balanced architecture based on your read-write ratio, query patterns, and consistency requirements.

How can I achieve zero-downtime database migration?

Zero-downtime migration requires a progressive migration strategy. Common approaches include: running systems in parallel (old and new systems operating simultaneously with gradual traffic shift), trickle migration (incremental data synchronization), and blue-green deployments (maintaining two full environments and switching via DNS/load balancer). Key steps include: 1) using CDC (Change Data Capture) tools to keep data synchronized; 2) fully validating the migration process in a test environment; 3) performing phased canary/gradual traffic switches; 4) maintaining a comprehensive rollback plan and data snapshots. For large table migrations, use chunked migration to avoid long-running table locks. Database Architect will design a detailed migration plan for you, including time windows for each phase, verification checkpoints, and rollback trigger conditions.