ICESEE-Container

Container recipes and build definitions for the ICESEE software ecosystem.

This repository provides Docker and Apptainer/Singularity container definitions for building and running ICESEE and its dependencies in a portable, reproducible, and platform-aware way. The containers are intended for use across:

• Local development environments
• Cloud platforms
• High-Performance Computing (HPC) clusters

The goal of this repository is to make ICESEE deployment easier, more reproducible, and more consistent across diverse compute environments.

---

Overview

Scientific software stacks often depend on complex combinations of compilers, MPI implementations, numerical libraries, and domain-specific packages. Differences between laptops, cloud virtual machines, and HPC clusters can make installation and reproducibility difficult.

ICESEE-Container addresses this problem by maintaining container definitions that package ICESEE and its required dependencies in a controlled environment. These recipes help users:

• build reproducible environments for development and research,
• reduce dependency and configuration issues,
• simplify onboarding for new users and collaborators,
• support deployment across heterogeneous systems,
• enable portability between local, cloud, and HPC resources.

---

What this repository contains

This repository includes container recipes for:

• Docker
  • Suitable for local development, CI/CD, cloud deployment, and workstation use.
• Apptainer / Singularity
  • Suitable for HPC systems where Docker is unavailable or restricted.

Depending on the recipe, containers may include:

• ICESEE source and runtime environment
• Scientific libraries and compiler toolchains
• MPI-enabled builds
• PETSc and related numerical dependencies
• Optional workflow-specific tools used by ICESEE

---

Use cases

The containers in this repository are intended to support three main environments.

1. Local development

For developers and researchers working on laptops or workstations, Docker-based environments provide a clean and reproducible setup for building, testing, and iterating on ICESEE.

2. Cloud deployment

For cloud-based workflows, these recipes provide portable environments that can be used on virtual machines, container platforms, or batch systems.

3. HPC execution

For HPC systems, Apptainer/Singularity recipes enable containerized execution within cluster policies and scheduler-based environments, while maintaining compatibility with system MPI and performance-oriented libraries where appropriate.

---

Repository goals

The main goals of ICESEE-Container are:

• Reproducibility
  Ensure that ICESEE builds and runs in consistent environments across systems.

• Portability
  Support the same scientific workflow on local machines, cloud resources, and HPC clusters.

• Maintainability
  Keep container definitions organized, versioned, and easy to update as ICESEE evolves.

• Scalability
  Enable transition from development environments to production and large-scale HPC runs.

• Accessibility
  Lower the barrier to entry for new users who need a working ICESEE environment without manually installing a large dependency stack.

---

Repository structure

A suggested structure for this repository is:

ICESEE-Container/
├── spack-managed/
│   ├── firedrake-container/
│   ├── icepack-container/
│   ├── issm-container/
│   ├── icesee-container/
│   └── firedrake-icepack-issm-icesee-container/
├── raw-recipes/
│   ├── firedrake-container/
│   ├── icepack-container/
│   ├── issm-container/
│   ├── issm-icesee-container/
│   └── firedrake-icepack-container/
└── README.md