Giters

iot-salzburg / gpu-jupyter

GPU-Jupyter: Leverage the flexibility of Jupyterlab through the power of your NVIDIA GPU to run your code from Tensorflow and Pytorch in collaborative notebooks on the GPU.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Container not accessible from the network with Podman instead of Docker

itodorovic opened this issue · comments

commented

Issue Description

When running the container with podman instead of docker, container is not visible from other devices on the network.
This seems to be discussed in upstream project:
b-data/data-science-devcontainers#1
b-data/data-science-devcontainers#1 (comment)
I'm using CDI.

To Reproduce
Steps to reproduce the behavior:
Run with:
podman run --device nvidia.com/gpu=all -d -it -p 8848:8888 -v $(pwd)/data:/home/jovyan/work -e GRANT_SUDO=yes -e JUPYTER_ENABLE_LAB=yes -e JUPYTER_TOKEN=${JUPYTER_TOKEN} --user root --restart always --name gpu-jupyter cschranz/gpu-jupyter:v1.6_cuda-12.0_ubuntu-22.04

Environment

Operating System:
RHEL 8

NVIDIA GPU and CUDA version Details:

# nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Wed_Nov_22_10:17:15_PST_2023
Cuda compilation tools, release 12.3, V12.3.107
Build cuda_12.3.r12.3/compiler.33567101_0

# nvidia-smi
Mon Apr  1 12:25:55 2024
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.54.14              Driver Version: 550.54.14      CUDA Version: 12.4     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  Tesla P6                       Off |   00000000:13:00.0 Off |                    0 |
| N/A   34C    P0             26W /   90W |    8128MiB /  15360MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    0   N/A  N/A     77564      C   /server                                      8126MiB |
+-----------------------------------------------------------------------------------------+
commented

[...] container is not visible from other devices on the network

Can you curl -v http://127.0.0.1:8848 from the same device (i.e. the container host)?

If so, then port 8848 is blocked (firewall?) and this issue is not related to the container.