Skip to main content

API - How Tos

Aggregates / Rollups​

How to implement ?​

  • Database SQL has evolved to offer advanced aggregation and grouping abilities. Why not leverage it?
  • To get any type of aggregation on the data returned from your database and thru APIs, first create database views on the tables, with any/all table joins and groupings.
  • One can also join 2 views with each other for multi-level aggregations or transformations.
  • One can form multi-level view utilizing WITH blocks in SQL if supported by database being used.
  • Then import final view object into Builder Studio and generate code on it.
  • The APIs generated on such object will fullfill data aggregation needs of your business requirement.

Backend Templates​

info

Templates are available for Spring Java backend. Coming soon for Express Node.js backend. One can low-code customize to implement them.

  • Pease refer to section Backend Templates

  • Code Coverage - SonarQube, JaCoCo

    • Code Coverage Analysis

  • Extend

    • Identifiers, @GeneratedValue
    • Validations
    • DTO (Data Transfer Object)

  • Security

    • OAuth2
    • Keycloak
    • Social (Google, Github, ...)

  • Logging - ELK (Elasticsearch, Logstash, and Kibana)

  • Integrating - Kafka, Kafka Streams

  • Testing

    • System Testing - Performance, Load testing
    • TDD (Test Driven Development)
    • Smoke Testing
    • BDD (Behavior-Driven Development)

  • Migrate to Spring Boot 3 (Spring Boot 3.2.3 and Java 17 or Java 21)

How to Use Redis Cache​

Redis cache can store and retrieve data quickly to improve the performance of your APIs. It is primarily designed as an in-memory data store for high-performance, low-latency data access.

How to run Redis From Docker​

docker pull redis:7.0.6-alpine
docker run --name my-redis -p6379:6379 -d redis

How to Stop redis docker container:

docker stop my-redis

Enable Redis Cache in application.properties​

  • For dbrest REST APIs:

    • Edit emapi\app\dbrest\src\main\resources\application.properties

  • For dbgraphql GraphQL APIs:

    • Edit emapi\app\dbgraphql\src\main\resources\application.properties

Locate section as below and follow instructions to enable:

# ---- Disabled redis cache state -------------
spring.cache.type=simple
...

# ---- Enable redis cache
# ---- By commenting above and un-commenting below -------------
...

Event-Driven Microservices - Kafka​

Build Event-Driven Microservices applications using Spring Boot and Apache Kafka.

Available with EasyManage Templates : Backend Templates : Kafka

Kafka Streams​

Kafka Streams enables the processing of streaming events.

Available with EasyManage Templates : Backend Templates : Kafka Streams

Distributed Transactions in Microservices​

How to manage distributed transactions across multiple microservices and also solve the problem of distributed locking ?

Distributed transaction in microservices means a local transaction in the monolithic system is distributed into multiple services that will be called in a sequence.

Possible solutions​

The following two patterns can resolve the problem:

  • Two-phase commit - 2pc
  • Saga Pattern

Two-phase commit - 2pc​

Two-phase commit Is used in database systems, not so much good fit for microservices.

  • In microservices, it is implemented as per below.
  • Uses two phases - A prepare phase and a commit phase
  • With a Global Coordinator across microservices.
  • Gives strong consistency and guarantee that the transaction is atomic.
  • But is synchronous (blocking), need to lock the object till transaction completes.

Saga Pattern​

In Saga Pattern

  • The distributed transaction is fulfilled by asynchronous local transactions on all related individual microservices. And finally completed.
  • The microservices communicate with each other through an event bus.
  • Each microservice fulfills its own local atomic transaction, other microservices are not blocked and there is no lock (kept throughout transaction) on any object.
  • In cae of failure of one operation within a sequence on a microservice, will results in all prior transactions being reversed using compensating strategy.

Notes:

  • Challenges: Can be diffcult to debug and maintain.
  • Add a process manager as an orchestrator.
  • Please refer to earlier section on Event-Driven Microservices - Kafka to implement such event based transaction processing.

Serverless​

  • Available with Spring Java Backend

Enabling Serverless

  • Go Serverless (Functions) with Spring Cloud Function

Spring Cloud Function​

WHY ? / HOW ?

Spring Cloud Function - Goals: 

- Promote the implementation of business logic via functions. 
- Decouple the development lifecycle of business logic from any specific runtime target 
    so that the same code can run as a web endpoint, a stream processor, or a task. 

Spring Cloud Function provides the following features: 
1. Wrappers for @Beans of type Function, Consumer and Supplier, 
exposing them to the outside world as either HTTP endpoints 
and/or message stream listeners/publishers with RabbitMQ, Kafka etc. 

Spring Cloud Function embraces and builds on top of the 3 core java functional interfaces: 
Supplier<O> 
Function<I, O> 
Consumer<I>

Please refer to Spring docs: Spring Cloud Function

How To Use

  • The generated code provides Templates For: Spring Cloud Function, e.g.
emapi\lib\base-app\src\main\java\com\example\emapi\app\ErpCustomer\ErpCustomerServiceCloudFunctions.java
  • These can be Integrated Further with any target Cloud: AWS, GCP, or Azure

Leverage them and Trigger via AWS Lambda, or deploy Spring Cloud Functions on any of AWS, Azure or Google Cloud.

AWS Lambda​

Please refer to Spring docs:

Microsoft Azure Functions​

Please refer to Spring docs: Microsoft Azure Functions

Google Cloud Functions​

Please refer to Spring docs: Google Cloud Functions