Nvds_msgapi_do_work Unable to publish, error-7

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU): Jetson Nano
• DeepStream Version: 5.0 GA
• JetPack Version (valid for Jetson only): 4.4 (L4T 32.4.3)
• Issue Type( questions, new requirements, bugs): Sudden error when publishing MQTT messages (AWS IoT Core)
• How to reproduce the issue ?: I have not been able to identify conditions to reproduce the error.

Hi everyone. I have a deepstream application running on a jetson nano in a “remote” location that I can access via VPN. This application uses PeopleNet to detect people and together with nvtracker and nvdsanalytics, it counts people who pass through a line. Finally send these results through MQTT (AWS IoT Core). After running for about a month without any inconvenience (only some manual restart of the application due to a configuration change), I started receiving these messages last Friday (Oct 16).

error_deepstream_mqtt
error_deepstream_mqtt1916×195 150 KB

Upon reviewing the application, it is still running and continues to perform detections and counts, however it is not able to publish MQTT messages. By stopping and restarting the application it begins to publish messages without problems again. However, since that day, the problem has been repeated three times without following any apparent pattern.

I am not an expert on the subject, but it strikes me that I suddenly started having this problem without having made any changes to the application. Could it be because of an external factor? Has anyone had a similar problem? Is there a way to handle this error so that the connection can be restarted automatically?

As additional information, mention that I followed the steps of aws-iot-core-integration-with-nvidia-deepstream to publish data to AWS IoT Core and that the payload sent is customized.

3

HI,
What kind of configuration change you made to restart the application and caused the error? are you sure this config change cause you receiving the “unable to publish error”?
and we do have support reconnection, in case remote broker have some failure like down or something like that, after the broker back, the application have retry policy to reconnect.
and please try to figure out how we can repro your issue?

6

Hello, thanks for the answer. Performing some tests with another jetson nano in the office, I was able to identify that if the system loses connection to the Internet for a period of 15 minutes, these error messages begin to appear and when reconnecting the problem remains.

On the other hand, if I disconnect the system from the internet for a period of less than 15 minutes (10 minutes for example), the system is able to dump all the outgoing messages generated while there was no connection without any problem.

7

I will try to repro the issue you mentioned, thanks for your patience.

1 Like
8

HI,
Sorry for the late.
I reproed the issue can not publish the message after service back, i shutdown amqp broker service to simulate your case out of network, after 20 minutes, it can not send message.
We are checking internally, and get back to you once have progress.
thanks.

9

Hi,
For now, auto reconnect doesnt work in any of DS message adapters.

1 Like