Running a docker image but its throwing an error

Docker for AWS
#1

I have created a docker image and getting the following error while running on a aws machine.

Could not load library libcudnn_ops_infer.so.8. Error: libcublas.so.11: cannot open shared object file: No such file or directory
Please make sure libcudnn_ops_infer.so.8 is in your library path!

As a solution I had inserted the following line in the docker file:
RUN export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-11.0/targets/x86_64-linux/lib/libcudnn_ops_infer.so.8

But am still getting the same error. If anyone could help me it would be great.

#2

Export will not help in the image. That variable will only exist in that layer created by the RUN instruction which does not make sense. Try this:

ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-11.0/targets/x86_64-linux/lib/libcudnn_ops_infer.so.8

Also make sure that the path exists and you installed the necessary libraries

#3

“installing the necessary libraries” means…, insertion of following lines into the docker file will do ?

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin
sudo mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600
sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/${last_public_key}.pub
sudo add-apt-repository "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/ /"
sudo apt-get update
sudo apt-get install libcudnn8
sudo apt-get install libcudnn8-dev
#4

If I knew what were the necessary libraries, I would have just told you that :slight_smile: The only thing I saw was the error message. If you those files are not even there, setting environment variables can’t help. Since I rarely use GPU with Docker and I have never seen that error message, this is all I could tell. You wrote days ago. I hope you could find a solution since then.