test-queue | parallel test runner for CI environments | Continous Integration library

As per the documentation, you can set/change the default region of the cloud functions from us-central1 through the below code example (I don’t know why you had put

regionalFunctions.runWith(expressOpts).https.onRequest(app);

It should be :

For some reason, the message type you are producing doesn't match the message type consumed by the consumer:

You don't have to include the Authorization header in the request. Simply use your SAS URL to send the request (with other headers) and the request should work fine.

The reason you do not need Authorization header is because the authorization information is already included in your SAS Token (sig portion of your SAS Token).

Your code would be something like:

At first I was confused by why, if test-queue.fifo is what is triggering the lambda function, you wouldn't just process event directly rather than polling the queue in the function logic. Thinking about it though, I think this pattern is your problem. Namely, here's what is happening:

1. A message arrives in the SQS queue.
2. The message is sent to your lambda and triggers its execution. It is no longer visible in the queue, because:

When Lambda reads a batch, the messages stay in the queue but are hidden for the length of the queue's visibility timeout. If your function successfully processes the batch, Lambda deletes the messages from the queue. (AWS docs source)

1. The lambda polls the queue. There's nothing (visible) there. The lambda returns successfully.
2. Because the lambda finished successfully, the event is deleted from the queue.

You should use the already-implemented-for-you wiring between Lambda and SQS. The link above is an overview of how to do this. You can choose how many messages should be sent (at most) at a time to your lambda, and you'll get them passed to the function, and automatically removed from the queue.

If you really really really want to use your manual approach, you'll need to either change the trigger of your lambda (the idea that comes to mind is a Cloudwatch metric trigger on the number of pending messages), or change the message visibility timeout and add logic to delete events from the queue after processing and make sure your polling + processing + deleting logic plays nicely with Lambda's auto-deletion-on-success. I do not recommend this latter approach :)

I think that you have to return from createInterval function the intervalId and after that it should work.

Can you check what value has your intervalId right now, with your current implementation?

https://developer.mozilla.org/en-US/docs/Web/API/setInterval

"The returned intervalID is a numeric, non-zero value which identifies the timer created by the call to setInterval(); this value can be passed to clearInterval() to cancel the interval."

So the issue is with service bus itself. We are using service bus in standard tier which doesn't support selector. More details here and here

The above issue is fixed after cleaning up connect-cluster-configs, connect-cluster-status, connect-cluster-offsets topics using following command in Kafka broker.

I think, if you change your acknowledge mode, the problem will be fixed: change your consumer part like this:

As it turns out i was looking for a proper use of function "log_()" and "get_selected()"