Memory Leak in DeepStream Python bindings

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) Jetson Xavier (AGX and NX)
• DeepStream Version 6.0
• JetPack Version (valid for Jetson only) 4.6
• Issue Type( questions, new requirements, bugs) Bugs

Hi, I have noticed that when running DeepStream apps in Python on Jetson Xavier devices for a long duration, the app seems to run out of memory at some point (gives an ‘OOM error’) and shut down. Not sure if this has an impact, but the app is run with a fakesink and in a Docker container, the base image of which is pulled from deepstream-l4t:6.0-samples

I have tested to see which components of the pipeline seem to be the culprit(s) and have narrowed it down to the PGIE. Two main observations lead me to believe this:

a) increasing the interval of the pgie from 0 to 1 allows the app run for double the amount of frames as before (~500k frames vs ~1m frames) until OOM.
b) when I unlink the PGIE from the pipeline the app is able to run indefinitely.

Is there a known fix for the memory leak or is it perhaps fixed in one of the newer versions of DeepStream?

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

Can you please retest it on the latest version? There are some bug fixing related with memory leak in the latest version.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.