Tensorrt: Installing TensorRT on docker | Depends: libnvinfer7 (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed

Created on 22 Sep 2020  路  7Comments  路  Source: NVIDIA/TensorRT

Description

My base system is ubuntu 18.04 with nvidia-driver. I just installed the driver and it is showing cuda 11.
While installing TensorRT in the docker it is showing me this error. I haven't installed any drivers in the docker image.

root@b2f8be4565fe:/home/asd/Downloads/docker_folder# apt-get install tensorrt
Reading package lists... Done
Building dependency tree       
Reading state information... Done
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
 tensorrt : Depends: libnvinfer7 (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed
            Depends: libnvinfer-plugin7 (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed
            Depends: libnvparsers7 (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed
            Depends: libnvonnxparsers7 (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed
            Depends: libnvinfer-bin (= 7.1.3-1+cuda10.2) but it is not going to be installed
            Depends: libnvinfer-dev (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed
            Depends: libnvinfer-plugin-dev (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed
            Depends: libnvparsers-dev (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed
            Depends: libnvonnxparsers-dev (= 7.1.3-1+cuda10.2) but 7.2.0-1+cuda11.0 is to be installed
            Depends: libnvinfer-samples (= 7.1.3-1+cuda10.2) but it is not going to be installed
            Depends: libnvinfer-doc (= 7.1.3-1+cuda10.2) but it is not going to be installed
E: Unable to correct problems, you have held broken packages.

NVIDIA-SMI 450.66 Driver Version: 450.66 CUDA Version: 11.0

Details about the docker
docker pull nvidia/cuda:10.2-devel-ubuntu18.04
dpkg -i libcudnn8_8.0.3.33-1+cuda10.2_amd64.deb
dpkg -i libcudnn8-dev_8.0.3.33-1+cuda10.2_amd64.deb

Environment

TensorRT Version: 7.1.3
GPU Type: 1050 TI
Nvidia Driver Version: 450.66
CUDA Version: 10.2
CUDNN Version: 8.0.3
Operating System + Version: Ubuntu 18.04
Python Version (if applicable): N/Aa
TensorFlow Version (if applicable): N/A
PyTorch Version (if applicable): N/
Baremetal or Container (which commit + image + tag): N/A

Steps To Reproduce

apt-get install tensorrt

DockeContainers triaged

Most helpful comment

Can you provide support Nvidia ?
We are stuck on our deployment for a very important client of ours.
TensorRT seems to taking cuda versions from the base machine instead of the docker for which it is installed.

All 7 comments

Can you provide support Nvidia ?
We are stuck on our deployment for a very important client of ours.
TensorRT seems to taking cuda versions from the base machine instead of the docker for which it is installed.

I am also experiencing this issue. Trying to get deepstream 5 and TensorRT 7.1.3.4 in a docker container and I came across this issue.

@tamisalex were you able to build this system?
Deepstream + TRT 7.1?

I abandoned trying to install inside a docker container.

I found that the CUDA docker image have an additional PPA repo registered /etc/apt/sources.list.d/nvidia-ml.list. This seems to overshadow the specific file deb repo with the cuda11.0 version of libnvinfer7.

I just added a line to delete nvidia-ml.list and it seems to install TensorRT 7.0 on CUDA 10.0 fine. I am not sure on the long term effects though, as my native Ubuntu install does not have nvidia-ml.list anyway.

This is documented on the official TensorRT docs page.
I was able to follow these instructions to install TensorRT 7.1.3 in the cuda10.2 container in @ashuezy 's original post. You would probably only need steps 2 and 4 since you're already using a CUDA container:

https://docs.nvidia.com/deeplearning/tensorrt/install-guide/index.html#maclearn-net-repo-install-rpm

We have the same problem as well.

  • Cuda 11.0.2
  • Cudnn 8.0
  • TensorRT 7.2

The following packages have unmet dependencies:
tensorrt : Depends: libnvinfer7 (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvinfer-plugin7 (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvparsers7 (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvonnxparsers7 (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvinfer-bin (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvinfer-dev (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvinfer-plugin-dev (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvparsers-dev (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvonnxparsers-dev (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvinfer-samples (= 7.2.2-1+cuda11.1) but it is not going to be installed
Depends: libnvinfer-doc (= 7.2.2-1+cuda11.1) but it is not going to be installed

Was this page helpful?
0 / 5 - 0 ratings

Related issues

WangXuanBT picture WangXuanBT  路  3Comments

mhansinger picture mhansinger  路  4Comments

prathik-naidu picture prathik-naidu  路  3Comments

dhkim0225 picture dhkim0225  路  6Comments

sbbug picture sbbug  路  5Comments