Somewhat related: #166

I submitted a PR to support a per-queue maxConcurrency field. However I'm wondering if this is still too inflexible. Perhaps a better approach would be to abstract the polling logic into a separate interface with a default implementation that could be overridden by users of the framework.

Hi! This PR is really very good. Thanks. I'm just waiting to have it in my project - waiting for release 🔢 .

I would like to give some suggestions based on my personal feeling when working with spring-cloud-aws, especially with SQS. Indeed I miss the maxConcurrency feature and it will be great to have it ASAP.
Beside that I think (and maybe I'm wrong) that also it would be great to have the following features as well:

1. Can we have also an additional attribute in SqsListener like maxNumberOfMessages?
   For example, if you have 2 queues, one 'my-queue-1' of type = standard queue and 'my-queue-2' of type FIFO queue. You would like to grab as many messages as possible from 'my-queue-1', and will want to preserve order during message processing on 'my-queue-2' - this means you will not use batching (one message at a time).
   Proposal: SqsListener#maxNumberOfMessages may be optional and when specified it will override
   SimpleMessageListenerContainer#setMaxNumberOfMessages().
   This will permit to configure the number of messages in batching for each queue. In example above, for 'my-queue-1' it ca be 10, but for 'my-queue-2' it can be 1;

2. Also it would be great to have a possibility to configure more than one consumer per queue. For example, if we have a FIFO queue with 50 groups, we can have 5 consumers that consume one message at a time. I know you will say that 'you can have one consumer that brings messages from SQS and passes them to another thread then continue to get next messages from SQS', still it would be nice to have such a feature - this will permit app scalability. Now we have only one consumer per queue , that brings messages sequentially (when using non-batching).
   Proposal: SqsListener#numberOfConsumers may be optional (default 1) and when specified with N value then N instances of AsynchronousMessageListener will be created. Or this attribute can have a value that will specify the core number of consumers and max consumers (like in JMS listener container), something like "2-10".

Sorry if this message is in wrong place.
Regards...

Hi @arseniir. I would like to ask you about your suggestion, because I have a situation that I have not been able to resolve, and I don't know if it is the same. I want to implement concurrency in the SqsListeners.

I read an AWS documentation about that, but it's a bit old, so I don't know if I need to change my implementation, or if there is a way to set the concurrency attributes in the config class.

Link: https://aws.amazon.com/blogs/developer/using-amazon-sqs-with-spring-boot-and-spring-jms/

The following is the implementation that I currently have, but I don't know how to set the concurrency attributes

Config class:

@Configuration
public class SqsConfig {

	@Bean
	public QueueMessagingTemplate queueMessagingTemplate(AmazonSQSAsync amazonSQSAsync) {
		return new QueueMessagingTemplate(amazonSQSAsync);
	}
	
	@Bean
	public AmazonSQS sqsClient() {
		return AmazonSQSClientBuilder.defaultClient();
	}
}

Consumer class:

@Service
public class Consumer implements ISqsStandardConsumer{

	@Override
	@SqsListener("SQS-concurrency-test")
	public void receiveStandardMessage(String message) {
		System.out.println(message);
	}
	
}

Do you know if there is already a way to do this?
Regards.

@juanledesma84 linked article doesn't use Spring Cloud AWS but amazon-sqs-java-messaging-lib

@maciejwalkowiak Doesn't really seem like awspring/spring-cloud-aws#40 addresses this. Seems like awspring/spring-cloud-aws#23 is probably the continuation of this?