Database in Microservices - (2nd PostgreSQL Conference Nepal 2023)

Practices for Database Design in Microservices
Architecture
Database in Microservices
using PostgreSQL
Sandip Basnet

About me
● Backend Developer.
● Using PostgreSQL for the last 4 years.
● Designing and building scalable and robust backend solutions.
● Believe in identifying issues and finding creative solutions that deliver tangible
results.
● Contact:
○ Email: mondyfy@gmail.com
○ LinkedIn: https://www.linkedin.com/in/sandipbasnet

Introduction
- Microservices architecture is an approach to building large, complex systems as a
collection of small, independent services.
- Each microservice has its own responsibility and can be developed, deployed, and
scaled independently.
- Databases are a critical component of microservices architecture, as they store and
manage data that is used by multiple services.

Problems
● Designing databases for microservices architecture is challenging because there
are many different ways to organize data and manage transactions.
● Some common challenges include:
○ Preserving ACID
○ Managing data consistency across multiple services.
■ Strong
■ Eventual
○ Ensuring fault tolerance.
○ Maintaining scalability as the system grows.

Approaches
● Architecture:
○ Multi-schema
● Maintain Isolation:
○ CQRS
● Manage transactions:
○ SAGA

Shared database/schema
● Services share the same database instance.
● Reduces the complexity of managing multiple databases, but it can also lead to
coupling between services.
● To avoid coupling, it's important to use clear boundaries and APIs between
services that share the same database.

Database per service
● Each microservice having its own dedicated database.
● Greater scalability, fault tolerance, and isolation between services.
● However, it can lead to increased complexity in managing and coordinating data
across multiple services.

Schema per service
● Each microservice has its own database schema.
● Different from the shared schema -> multiple microservices share the same
database schema.
● Provides improved scalability, easy maintenance, and better data isolation
between services.

CQRS (Command Query Responsibility Segregation)
● Separates read and write
operations into different
components.
● Improves performance and
scalability by allowing each
component to scale
independently.
● Microservices can implement
CQRS by using different
databases or database schemas
for reads and writes.

Saga Pattern
● Way to manage data consistency across microservices in distributed transaction
scenarios
● There are two types of patterns:
○ Choreography - each microservice publishes events to trigger the next step in the transaction.
○ Orchestration - a central component coordinates the transaction between microservices.

Saga - Choreographer
● Each microservice knows its
own part in the transaction and
publishes events to trigger the
next step.
● Provides greater autonomy and
scalability for each microservice.
● However, it can be more
difficult to coordinate and
manage transactions across
multiple services.

Saga - Orchestrator
● A central component (the
orchestrator) coordinates the
transaction between microservices.
● Provides greater control and
consistency over the transaction.
● However, it can lead to increased
complexity and coupling between
services.

Designing a microservices database schema
● When designing a microservices database schema with PostgreSQL, it's important
to consider the data access patterns and the relationships between different
services.
● Some key design principles include normalizing the data to minimize duplication,
using foreign keys to enforce relationships, and using indexing to optimize
performance.
● It's also important to consider the impact of schema changes on other
microservices that may depend on the data.

Challenges and considerations
● Managing data consistency across multiple databases,
● Implementing proper security measures, and
● Ensuring proper monitoring and management of the databases.
● Cost
● Scalability

Practices for using PostgreSQL in microservices
● Use a separate database/schema for each microservice to ensure isolation and
scalability.
● Implement CQRS to separate read and write operations and improve
performance.
● Use the SAGA pattern to manage data consistency across multiple microservices.
● Use JSON data types to handle complex data structures.
● Use PostgreSQL's support for ACID transactions to ensure data consistency and
reliability.

Why PostgreSQL for microservices
● ACID compliance.
● Robust feature sets
○ Support for JSON data types.
○ Ability to handle complex queries.
○ Object-oriented
● SQL Standard
● Open source development & Community

Summary
● Designing a database for microservices architecture using PostgreSQL can be a
complex task, but by following best practices and design principles, it can lead to
a robust and scalable system.
● By using separate databases for each microservice, implementing CQRS and
SAGA patterns, and using PostgreSQL's advanced features, developers can build a
reliable and scalable microservices.

Thank you for your time and attention!
I hope this session has been informative and helpful.
@mondyfy

References
● PostgreSQL as a Microservice: https://tapoueh.org/blog/2021/06/postgresql-as-a-microservice/
● Monolith to Microservices: Refactoring Relational Database
https://programmaticponderings.com/2022/04/14/monolith-to-microservices-refactoring-relational-
databases%EF%BF%BC/
● PostgreSQL in a microservice World: https://momjian.us/main/writings/pgsql/microservices.pdf
● Microservice Architecture: https://github.com/codefarm0/Microservices-Architecture
● Benefits: https://www.prisma.io/dataguide/postgresql/benefits-of-postgresql#object-oriented-database-
features

Database in Microservices - (2nd PostgreSQL Conference Nepal 2023)

More Related Content

What's hot (20)

Similar to Database in Microservices - (2nd PostgreSQL Conference Nepal 2023) (20)

More from Sandip Basnet (7)

Recently uploaded (20)

Database in Microservices - (2nd PostgreSQL Conference Nepal 2023)