Allow default Queue visibility timeout parameters to be customized

Currently in our default QueueProcessor we hard code the visibility timeout for failed queue trigger invocations to TimeSpan.Zero (code here). While you can write and register your own custom queue processor to override this, we should provide a JobHostQueuesConfiguration knob at the host level that lets this default be set globally.

A configuration knob is particularly IMPORTANT in scenarios like Azure Functions where the possibility for users to write a custom QueueProcessor doesn’t exist.

[dnlmcv]

Hi @mathewc ,

Still it is not working with azure functions. I’ve created an azure functions project in Visual Studio and configured the host.json file with following values: { “queues”: { “visibilityTimeout” : “00:01:00” } } The messages are retried with 0 seconds timeout. Any input on this ?

Thank you

[dnlmcv]

Will the visibility timeout only apply to retried queue items or does it apply to newly added items to the queue as well?

I don’t think so.

I believe your best chance of achieving some kind of transient fault handling is to create your own implementation where you do following:

• Clone the brokered message
• Add custom property for the number of retries
• Add an exponential or logarithmic logic for setting the enqueue time.
• Schedule the message and add it to queue for an execution at a later time.

I hope this helps.