No GPU temperature when running jtop inside a docker container
livneros-greeneye opened this issue · comments
Describe the bug
While running jtop inside a docker container, the GPU temperature details are offline.
To Reproduce
Steps to reproduce the behavior:
from jtop import jtop
jetson = jtop()
jetson.start()
jetson.temperature["GPU"]
- Running the following commands on my host plots:
> {'temp': -256.0, 'online': False}
- Running a docker container, as mentioned in the documentation:
docker run --rm -it -v /run/jtop.sock:/run/jtop.sock rbonghi/jetson_stats:latest
plots:
> {'temp': 39.875, 'online': True}
The same issue happens while using jtop
:
Board
Software part of jetson-stats 4.2.2 - (c) 2023, Raffaello Bonghi
Model: Jetson AGX Orin as NX-16GB - Jetpack 5.1.1 [L4T 35.3.1]
NV Power Mode[0]: MAXN
Serial Number: [XXX Show with: jetson_release -s XXX]
Hardware:
- 699-level Part Number: 699-13701-0005-500 M.0
- P-Number: p3701-0005
- Module: NVIDIA Jetson AGX Orin (64GB ram)
- SoC: tegra23x
- CUDA Arch BIN: 8.7
- Codename: Concord
Platform:
- Machine: aarch64
- System: Linux
- Distribution: Ubuntu 20.04 focal
- Release: 5.10.104-tegra
- Python: 3.8.10
jtop:
- Version: 4.2.2
- Service: Active
Libraries:
- CUDA: 11.4.315
- cuDNN: 8.6.0.166
- TensorRT: 5.1.1
- VPI: 2.2.7
- Vulkan: 1.3.204
- OpenCV: 4.5.4 - with CUDA: NO
I apologize for this huge delay in replying.
The GPU output you read from jtop is exactly what you notice from the NVIDIA output in devices/virtual/thermal.
With Jetpack 5 and Jetpack 6, when the GPU is not working also, the GPU output is equivalent to -256, which in jtop is "Offline".
The GPU output on your host and on your docker must be the same. The Jtop service reads this output one time and provides for every Jtop client the same value.
I updated jtop for the latest version of Jetpack 6.0
sudo pip3 install -U jetson-stats
I am closing this issue. If you encounter any new issues related to Docker, please feel free to reopen this issue.