• Hardware Platform (Jetson / GPU)
RTX 4090
• DeepStream Version
6.2
• TensorRT Version
8.5.2-1
• NVIDIA GPU Driver Version (valid for GPU only)
520.61.05
• Issue Type( questions, new requirements, bugs)
I am using the nvidia docker image nvcr.io/nvidia/deepstream:6.2-devel , trying to use the nvmsgconv-element in my pipeline.
I added some custom handling of my model output data to the nvmsgconv-library in order to send my desired data to kafka, after which I recompile the nvmsgconv-lib and use it instead of the original one.
When running the pipeline using my adjusted message conversion library, it works on some other machines (using RTX 3090’s although I cannot see how this would have an impact on nvmsgconv…?). However, the machine I’ve tried running it on (with the RTX 4090) gives the following error:
“One element could not be created. Exiting.”
When running the application with GST_DEBUG=5, I get the following information from gstelementfactory :
no such element factory "nvmsgconv"
When using the original message conversion library instead on the machine getting the error (and therefore not getting my desired output to kafka), it works to run the pipeline.
I assume there is some change within my version of the nvmsgconv-library causing this issue since the msgconv-element is created correctly when using the original version of the library, but I don’t understand what might be causing the issue ( since the errors occurs already when trying to create the nvmsgconv-element) and why it only occurs on some machines. I could really use some insight as of how to debug this.
( Side note: same problem was mentioned but not resolved in: GST_ELEMENT_FACTORY gstelementfactory.c:458:gst_element_factory_make: no such element factory "nvmsgconv"! , otherwise I havn’t been able to find much information about this issue )