Software Management on REPACSS¶
This section covers all aspects of software management on REPACSS, organized by package management approach and use case.
Software Management Categories¶
1. HPC Applications and System Software (Spack and Module System)¶
System-provided software - Users simply load what they need.
For HPC applications, compilers, libraries, and system-level software managed through Spack and loaded via the Lmod environment module system. No installation required - just load the modules you need.
Start here: Module System
2. Data Science Applications (Conda/MiniForge)¶
User-installed software - Users install locally, then add packages.
For Python/R-based data science, machine learning, and scientific computing packages using MiniForge (conda) environments. Requires local installation of MiniForge, then install packages like Jupyter Notebook for interactive development.
Start here: Getting Started with MiniForge
3. User Self-Installed Software (Custom Solutions)¶
User-installed software - Users install their own software using:
3.1 Containers with Apptainer (Pre-built Binaries)¶
For pre-built applications, complex software stacks, and reproducible environments. Examples include Ollama for AI applications.
Start here: Using Containers
3.2 Building from Source (Custom Compilation)¶
For software that needs to be compiled from source code with specific configurations.
Start here: Building from Source
Popular User Applications¶
Commonly Used Software¶
Popular applications that users frequently install and use, demonstrating how to use the package management methods above:
- Running Jupyter Notebook - Using conda (category 2) for interactive development
- Running Ollama - Using containers (category 3.1) for AI applications
Quick Reference¶
| Category | Installation Type | Use Case | Solution | Documentation |
|---|---|---|---|---|
| HPC Applications | System-provided | Compilers, MPI, scientific libraries | Module system | Module System |
| Data Science | User-installed | Python/R packages, ML frameworks | MiniForge (conda) | Getting Started with MiniForge |
| Pre-built Applications | User-installed | Complex software stacks | Apptainer containers | Using Containers |
| Custom Software | User-installed | Source compilation, specialized tools | Build from source | Building from Source |
Popular User Applications¶
| Application | Package Management Method | Installation Type | Documentation |
|---|---|---|---|
| Jupyter Notebook | Conda (Data Science) | User-installed | Running Jupyter Notebook |
| Ollama | Containers (User Self-Install) | User-installed | Running Ollama |
Getting Started¶
- HPC applications? Check available modules:
module avail(system-provided) - Data science work? Set up MiniForge: See Getting Started with MiniForge (user-installed)
- Custom software? Choose containers or source build based on your needs (user-installed)
- Want popular applications? See commonly used software: Popular User Applications
Documentation Structure¶
Package Management Methods¶
- Module System - System-provided HPC applications via Spack + Lmod
- Getting Started with MiniForge - User-installed data science with conda
- Using Containers - User-installed Apptainer/Singularity containers
- Building from Source - User-installed source compilation
Popular User Applications¶
- Running Jupyter Notebook - Using conda for interactive development
- Running Ollama - Using containers for AI applications
Need Help?¶
For software-related issues or to request new software installations, contact:
repacss.support@ttu.edu