Jetson Nano crash after 20-30 hours of running DS pipeline

• Hardware Platform (Jetson / GPU) Jetson Nano
• DeepStream Version 5.0
• JetPack Version (valid for Jetson only) 4.4-b144
• TensorRT Version 7.1.3
• Issue Type( questions, new requirements, bugs) question

Good evening,

I have been experiencing crashes of my Jetson Nano for couple of months now.
I am constantly running a customized version of deepstream-test3 in which I also have integrated the messaging part with AMQP.

After several hours (usually 20-30) the Jetson Nano does not respond anymore. It is not even possible to access to Jetson Nano to ssh. However, I can still ping the board and it responds back.

I have looked at the log, but there is nothing strange. Not in the syslog and not in the kern.log.
The only “strange” message that I get in the kern.log continuosly is the following:

Sep 14 07:23:28 jnano-desktop kernel: [92192.089364] nvmap_alloc_handle: PID 32514: deepstream-test: WARNING: All NvMap Allocations must have a tag to identify the subsystem allocating memory.Please pass the tag to the API call NvRmMemHanldeAllocAttr() or relevant. 
Sep 14 07:28:28 jnano-desktop kernel: [92492.125265] nvmap_alloc_handle: PID 32684: deepstream-test: WARNING: All NvMap Allocations must have a tag to identify the subsystem allocating 

Do you think that a bad memory allocation might be related to the issue?

Do you have any suggestion?

Thank you very much!!

Can you upgrade to DS 5.1?

Hi Kesong,

which is the correct way to upgrade it?
Should I just install the new SDK by installing it from the archive deepstream_sdk_v5.1.0_jetson.tbz2?

Thanks!

Quickstart Guide — DeepStream 5.1 Release documentation (nvidia.com)