Aws::SQS::QueuePoller automatically call change_message_visibility_timeout while work is in progress

[michaelwittig]

Is your feature request related to a problem? Please describe.
I use the Aws::SQS::QueuePoller to get messages from SQS. I can either define an :visibility_timeout or the queue's default value is used. I can also call poller.change_message_visibility_timeout(msg, 60) to add another timeout if I need more time. But in situations where the processing is an expensive blocking call, I can not easily do that (e.g., download a large file from S3).

Describe the solution you'd like
So I was asking myself, why can't the poller automatically call change_message_visibility_timeout every x seconds to reset the timeout x+y while the work is in progress (code has neither returned or thrown).

Describe alternatives you've considered
One idea I had is to start a Thread inside the block that calls the change_message_visibility_timeout from time to time until I kill it when I reached the end of my code block.

[ ] 👋 I may be able to implement this feature request

Additional context
Add any other context or screenshots about the feature request here.

[github-actions]

Greetings! We’re closing this issue because it has been open a long time and hasn’t been updated in a while and may not be getting the attention it deserves. We encourage you to check if this is still an issue in the latest release and if you find that this is still a problem, please feel free to comment or open a new issue.

[JohnLegrandRichards]

I would be willing to look into this @mullermp. Are there any issues with adding functionality to these generated files?

[mullermp]

Thanks for reaching out about this. QueuePoller is not code generated, so you are free to change the file. As far as the feature request goes, on the surface, having a managed thread that updates the timeout until the block has returned might be ok, but it definitely needs to be an opt-in feature. I don't like the possibility that the poller could be blocked infinitely with no timeout - we still should configure a maximum. AFAIK the maximum visibility timeout is already 12 hours - are there really s3 downloads that take that long? I am doubting the need for this feature request. @JohnLegrandRichards are you by chance running into an issue in your application that requires this feature?

[mullermp]

I am closing this because 12 hours is an extremely generous maximum visibility timeout. We will re-consider with more customer data.

[github-actions]

⚠️COMMENT VISIBILITY WARNING⚠️

Comments on closed issues are hard for our team to see.
If you need more assistance, please either tag a team member or open a new issue that references this one.
If you wish to keep having a conversation with other community members under this issue feel free to do so.

[michaelwittig]

@mullermp the point is not that something takes longer than 12 hours.

The point is that you can consistently tell SQS that the message is still worked on. In case a worker dies, the messages is quickly picked up by another worker. Instead of waiting for the 12 hours to end (or whatever the visibility timeout was).

[mullermp]

That makes sense, sorry, I will re-open this.

[RanVaknin]

Hi there,

After reviewing this FR with the team once again we have decided not to move forward with implementing this. This will not be an easy thing to implement and the team is prioritizing other features at this time. If this comes up again in the future we will likely prioritize this as a cross SDK feature.

Thanks again for taking the time and submitting the request.

All the best,
Ran~

[github-actions]

This issue is now closed. Comments on closed issues are hard for our team to see.
If you need more assistance, please open a new issue that references this one.