I hava been reading about using pollTimeoutSeconds in @SqsListener from Spring Cloud AWS.
Look at this configuration:
@SqsListener(value = "my-queue", pollTimeoutSeconds = "10", maxMessagesPerPoll = "1", maxConcurrentMessages = "1")
Reading the documentation I understood that:
If a message arrives within 10 seconds, it's immediately processed. I did the test and confirmed this.
I understood that for every 10 seconds, a new pool will be started again. I could see that on CloudWatch Graphic.
The only difference is that using 1s a request will be fired every 1s, with 10s the request will wait for 10s. (more requests, more costs)
My question is: in terms of receiving messages there is no difference between 10 or 1 second, messages will always be processed immediately when they become visible in SQS, is that correct?
messages will always be processed immediately when they become visible in SQS, is that correct?
Immediately is a strong word, but for simplicity, I would say yes.
If a client has started polling and a message enters the queue, SQS will send a response back to client with this message.
pollTimeoutSeconds does not affect response time.
If a message is available, the call returns sooner than pollTimeoutSeconds
pollTimeoutSeconds is a same thing as WaitTimeSeconds parameter of ReceiveMessage request. About them you could read here: https://docs.aws.amazon.com/AWSSimpleQueueService/latest/APIReference/API_ReceiveMessage.html