Verdin imx8m plus 1.1A - memory full - SoM restarting itself

Hello community,

We are using the Torizon SOM and Carrier Boards as mentioned below,

iMX8M Plus Q 4GB WBIT 16GB V1.1A
Customized Carrier Board
Torizoncore preempt RT - 5.7.0
toradex easy installer - 5.7.0

HOST PC - Ubuntu 22.04

We are using this specific version of the core to maintain the kernel compatibility with the other softwares.

The problem being faced is that the device is restating itself. I guess it is due to memory issue as we are logging our application data as bin files in our ROS environment inside the docker container

I’m attaching the screenshots
Uploading: Screenshot from 2024-10-24 12-03-30.png…

i have tried to stop the docker, and also remove the docker container, but before that command executes the connection is getting broken.

how to enter inside the docker container to delete or recover the log and make the device accessible?

Hi, I’m the Toradex AI Assistant. The Toradex Support team will follow up with you shortly!

Did you like my answer?

:+1: :-1:


Based on the context provided, it seems that the device is restarting due to memory constraints, likely caused by excessive logging within your ROS environment inside a Docker container. Here are the steps you can take to address this issue:

  1. Access the Docker Container: Use the command docker exec -it <container_id> /bin/bash to enter the Docker container. Replace <container_id> with the actual ID of your running container. This will allow you to access the shell of the container and manage files.

  2. Monitor Memory Usage: Use the command docker stats to monitor the memory usage of your Docker containers. This will help you identify if the memory usage is indeed the cause of the restarts.

  3. Optimize Logging: Implement log rotation or limit the size of log files generated by your application to prevent memory overflow. This can help in managing the memory usage more effectively.

  4. Alternative Access Methods: If you are unable to maintain a stable connection to execute commands, consider using a serial console or SSH to access the device directly. This might provide a more reliable connection for managing Docker containers and recovering logs.

  5. Recovery Mode: If the device is unresponsive, you may need to reboot it in a safe mode or use recovery tools to access the container and manage the logs.

By following these steps, you should be able to mitigate the memory issue and stabilize your device. If the problem persists, further investigation into your logging practices and memory management strategies may be necessary.

Please, also check if these links can help you solve your problem:

Memory Allocation Issue: Cortex M7 - m_data Insufficient Size.

Hello,

Can u please suggest how to reboot verdin in safe mode and what are the recovery tools available for doing the same?

Best Regards,
Nisha

Hi Nisha,
You can re-flash the module using Toradex easy installer

Hi Sahil,

Is there any approach other than reflashing?