nvidia-smi dosen’t work after flattening official nvidia docker image
I followed the directions listed here:
https://docs.nvidia.com/deeplearning/dgx/user-guide/index.html#port-flatten
Steps to reproduce:
$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
ubuntu 16.04 a51debf7e1eb 3 weeks ago 116MB
nvidia/cuda 10.0-devel-ubuntu16.04 e48eeb7fd785 4 weeks ago 2.26GB
$ nvidia-docker run --name nvidia nvidia/cuda:10.0-devel-ubuntu16.04 nvidia-smi
Thu Dec 13 21:35:00 2018
±----------------------------------------------------------------------------+
| NVIDIA-SMI 410.79 Driver Version: 410.79 CUDA Version: 10.0 |
|-------------------------------±---------------------±---------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce GTX 750 Ti Off | 00000000:08:00.0 On | N/A |
| 40% 31C P8 1W / 38W | 121MiB / 2000MiB | 30% Default |
±------------------------------±---------------------±---------------------+
| 1 Tesla P100-PCIE… Off | 00000000:84:00.0 Off | 0 |
| N/A 36C P0 31W / 250W | 0MiB / 16280MiB | 0% Default |
±------------------------------±---------------------±---------------------+
±----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
±----------------------------------------------------------------------------+
$ docker export nvidia | docker import - nvidia/cuda:flat
$ nvidia-docker run --rm nvidia/cuda:flat nvidia-smi
standard_init_linux.go:190: exec user process caused “exec format error”