This repository contains material for demonstrating, learning, and testing the use of the Trilinos.
The Trilinos Project is a community of developers, users and user-developers focused on collaborative creation of algorithms and enabling technologies within an object-oriented software framework for the solution of large-scale, complex multi-physics engineering and scientific problems on new and emerging high-performance computing (HPC) architectures.
In order to make the material
- easy to use,
- accessible, and
- portable,
the repository provides scripts to build a Docker container with all required software. Within the Docker container, it is easy to compile, run, and modify the exercises and tutorials for Trilinos and Kokkos on CPUs.
Note: Of course, the required software can also be installed directly, such that the examples can be compiled and run without the Docker container.
Docker can be easily installed on Linux, Windows, and MacOS and facilitates the delivery software in packages called containers, which are isolated and bundle their own software, libraries, and configuration files.
Please first install Docker on your computer following the instructions on the official website.
All following steps require that the software Docker is running on your computer. Moreover, if Docker complains about permissions, add your user to the docker group.
It is assumed that you have cloned this repository to your machine.
The Docker container can be built by executing the script
./build-container.shin the main directory of the repository. Running this script will set up a docker image with the flag trilinos_demo as described above. Since all the above mentioned software will be installed, this step will take some time (> 1 hour).
In order to run the Docker container, just execute the script
./run-container.shfrom the main directory of the repository. The script will run the container and mount the current directory (should be the main directory of the repository) as the local directory /opt/trilinos_demo within the container.
A pre-installed Trilinos is located at /opt/trilinos/ with its subdirectories build/, install/, and source/. The preinstalled Kokkos with OpenMP is located at /opt/kokkos/.
For the hands-on Trilinos exercises, navigate to /opt/trilinos_demo/exercises/ and follow the instructions in the respective README.md files.
For the hands-on kokkos exercises, navigate to /opt/kokkos-tutorials/ and follow the instructions in the respective README.md files.
To stop the containter, type exit in the container's command line.
In case you later want to remove the Docker image and the respective mounted volume (from running the container), just execute the script
./delete-container.shin the main directory of the repository.
-
EuroTUG
-
Trilinos
-
Kokkos
- When using a linux based system like Ubuntu, the Docker images and container might be stored in the root directory. If the root directory storage capacity is not sufficient please see this post for changing the directory to another location.