Getting Started

This will teach you how to get NumIO ensembles up and running.

Introduction

What is NumIO Ensembles?

Ensembles is a helper for the NumIO benchmarking tool. It aims to be quick to learn and easy to use.

When running applications in a cluster computing environment there are often other jobs running in the background. Ensembles attempts to replicate an environment like this to provide realistic benchmarking results.

Assumptions

This guide will assume you are usign a compute Cluster with Linux and MPI installed. It will also assume basic knowledge of these things. This guide might be able to help if you are not familiar yet.

Setup

Prerequisites

NumIO Ensembles depends on a few things. To follow this guide you will need a few things:

NumIO
access to a cluster computing environment with Linux, MPI and Slurm
the git CLI
Conda

Clone the Repository

NumIO ensembles is available on GitHub. After you have logged into your account on the cluster run:

git clone --filter=tree:0 https://github.com/fruitsbat/ensembles/

This will create a folder called "ensembles".

Install the python dependencies

Now cd into the folder:

cd ensembles

NumIO Ensembles needs a few python packages to work. A conda environment is provided for this. You can install it by running:

conda env create --name ensembles --file=environment.yml

Then load it with:

conda activate ensembles

Of course you don't need to name your environment "ensembles" for this to work.

Create a batch file

Now you will need to create a batch file for Slurm. Check the config section for more info.