NLU node stops working after 200 calls

[gxvigo]

Hi There, we built a flow that fetches customer feedbacks (text) from a DB and pass to NLU for processing. The DB has about 4000 records, NLU seems to stop processing records after about 200. We can't see any message in the Debug tab, so we can't really understand what kind of limit we are hitting.
Is there a way to get the cause of the issue?
We are using NodeRed on IBM Cloud.

[hansb001]

I am not sure this is the issue, but NLU works with so-called NLU Items, you have 30.000 NLU Items per month in the lite Plan. If you look in the catalog on IBM Cloud, and see the description of the NLU service, you can see how these NLU items are calculated.

[gxvigo]

I saw that, thanks. It doesn't explain the behaviour though. If we hit the 30.000 NLU we shouldn't be able to make further calls the end of the period (1 month), while restarting the process we can get few hundreds more. That's why I'd like to see the response code from the NLU HTTP call, I believe it could give some insights about the problem we are facing.

[chughts]

That is odd, because if the call is good, then the data response, no matter what it is is put on msg.features. I.e. Even if it says 'you have run out of credit' then it will be on msg.features.

if an error is returned from the service, it is reported in the status.

If you are seeing either then the service is not responding.

[gxvigo]

I will add a test on that parameter.
Where HTTP response codes (e.g 404, 401, 500, etc) are recorded?

[gxvigo]

Another thing I thought is about the number of API calls per period of time (seconds/minutes) that the NLU service can handle. I can't find any details in the documentation.