0

I am using resilience4j.retry with resilience4j.circuitbreaker. Service1 is calling another service Service2, which can throw exceptions. Even if I get exceptions, I should try to connect for at least defined no of times waitDuration * maxRetryAttempts.

Service1 ---> calling ---> Service2

application.yml 

resilience4j.retry:
  instances:
    service1:
      maxRetryAttempts: 4
      waitDuration: 1000 #ms
      retryExceptions:  
        - org.apache.http.conn.HttpHostConnectException
        - org.apache.thrift.TException
        - java.net.ConnectException

Service1Repository.java

Service2Response response = CircuitBreaker.decorateSupplier(circuitBreaker,
                                    Retry.decorateSupplier(retry, () -> {
                                        try {
                                            return service2.getdata(params);
                                        } catch (TException e) {
                                            log.error("msg", e);
                                            throw new Service1Exception("msg", e);
                                        }
                                    })
                            ).get();

However, when Service2 is down, it is not retrying (max 4 times, each after 1000ms should happen) I am getting a response (null in my case) immediately. I tried with increasing values (10000ms), still the same. I am seeing logs are printed but service should wait for waitDuration * maxRetryAttempts if Serviece2 is down.

roottraveller
  • 7,942
  • 7
  • 60
  • 65

1 Answers1

0

If you want to catch checked exceptions, you have to use decorateCallable.

Robert Winkler
  • 1,734
  • 9
  • 8