Resilience4j- Session-2 Limit of number of concurrent requests to RESTAPI using bulkhead

In this tutorial we are going to learn how to limit number of concurrent (parallel) requests that a restapi can handle.

Overview

• Using Bulkhead to limit the number of concurrent calls to a service.
• This functionality can be achieved easily with annotation @Bulkhead without writing explicit code.

Overview

• User makes a call to greeting RESTAPI to get a greeting message.
• greeting api can only handle can only handle 5 concurrent requests at a given time.
• This limitation can be set on greeting method using annotation @Bulkhead
• if the method receives more than 5 concurrent requests at a given time io.github.resilience4j.bulkhead.BulkheadFullException is thrown and user is served with HTTP status code 429 (Too Many Requests)

Source Code

• https://github.com/balajich/resilience4j-helloworld/tree/master/bulkhead

Video

• https://youtu.be/35fTH9G8gT0

Architecture

Prerequisite

• User should have basic understanding of RESTAPI and Spring boot
• JDK 1.8 or above
• Apache Maven 3.6.3 or above

Build

• cd resilience4j-helloworld
• mvn clean install

Running RESTAPI

• REST API Server: java -jar .\bulkhead\target\bulkhead-0.0.1-SNAPSHOT.jar

Using JMeter to generate more than 5 concurrent calls

• JMeter Script is provided to generate calls
• Import resilience4j-helloworld.jmx and run bulkhead thread group.
• Obeseve only 5 concurrent calls are successful rest of them fail with 503

Code

Include following artifacts as dependency for spring boot restapi application. resilience4j-spring-boot2, spring-boot-starter-actuator,spring-boot-starter-aop pom.xml of bulkhead

<dependency>
    <groupId>io.github.resilience4j</groupId>
    <artifactId>resilience4j-spring-boot2</artifactId>
    <version>1.4.0</version>
</dependency>
<dependency>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-actuator</artifactId>
</dependency>
<dependency>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-aop</artifactId>
</dependency>

In application.yml define the behavior of bulkhead module

• maxConcurrentCalls: Number of maximum concurrent calls that it should handle.
• maxWaitDuration: Duration for which request should wait.

resilience4j:
    bulkhead:
        configs:
            default:
                maxConcurrentCalls: 5
                maxWaitDuration: 0
        instances:
            greetingBulkhead:
                baseConfig: default
        @GetMapping("/greeting")
        @Bulkhead(name = "greetingBulkhead", fallbackMethod = "greetingFallBack")
        public ResponseEntity greeting(@RequestParam(value = "name", defaultValue = "World") String name) {
            return ResponseEntity.ok().body("Hello World: " + name);
        }
    
    
        public ResponseEntity greetingFallBack(String name, io.github.resilience4j.bulkhead.BulkheadFullException ex) {
            System.out.println("BulkHead applied no further calls are accepted");
            HttpHeaders responseHeaders = new HttpHeaders();
            responseHeaders.set("Retry-After", "10"); //retry after 10 seconds
    
            return ResponseEntity.status(HttpStatus.SERVICE_UNAVAILABLE)
                    .headers(responseHeaders) //send retry header
                    .body("Too many concurrent requests- Please try after some time");
        }

References

• https://developer.mozilla.org/en-US/docs/Web/HTTP/Status
• https://www.baeldung.com/resilience4j
• Hands-On Microservices with Spring Boot and Spring Cloud: Build and deploy Java microservices using Spring Cloud, Istio, and Kubernetes -Magnus Larsson

Next Tutorial

How to deploy microservices using docker

• https://github.com/balajich/spring-cloud-session-6-microservices-deployment-docker