Deployment of a Python application for Data Science Research Infrastructure for Neuroscience (DSRIN)
This code serves as a template for building data science applications on a computing cluster using OpenShift
and Kubernetes
. This template will perform a simple GLM on a some high-dimensional toy data in parallel using multiple cluster threads (cpus
) in parallel.
- Introduce
Docker
for containerization of applications. - Introduce
OpenShift
andKurbernetes
platforms for high-performance computing (hpc). - Introduce
MPI
for parallel processing. - Introduce
SnakeMake
as an workflow tool.
The Python code sample uses the 8081 HTTP port.
Before you begin creating an application with this devfile
code sample, it's helpful to understand the relationship between the devfile
and Dockerfile
and how they contribute to your build. You can find these files at the following URLs:
- The
devfile.yaml
file has animage-build
component that points to yourDockerfile
. - The
docker/Dockerfile
contains the instructions you need to build the code sample as a container image. - The
devfile.yaml
kubernetes-deploy
component points to adeploy.yaml
file that contains instructions for deploying the built container image. - The
devfile.yaml
deploy
command completes the outerloop deployment phase by pointing to theimage-build
andkubernetes-deploy
components to create your application.
- For more information about Python, see Python.
- For more information about devfiles, see Devfile.io.
- For more information about Dockerfiles, see Dockerfile reference.