Software Preservation and Sustainability

Strategic Rationale and Evidence Base

Strategic Importance

Software is as essential as data in enabling scientific discovery of data-intensive and compute-intensive domains such as HEP and RA. Software environments are becoming increasingly complex, tightly coupled to specific hardware, and subject to rapid technological change. Without long-term preservation strategies, scientific results cannot be reproduced, validated, or reused. SPECTRUM’s vision of a federated, cross-domain compute and data continuum depends on stable, portable, and durable software artefacts. Software preservation ensures continuity across successive generations of HPC/HTC/Cloud architectures and enables reproducible science at scale. It reduces technical debt, enhances operational efficiency, and supports scientific autonomy by maintaining Europe’s ability to re-run, verify, and extend past analyses

Current State

While some communities (e.g., HEP) have mature practices for preserving code, configurations, and containerised environments, the European landscape remains fragmented. This maturity difference is also reflected when comparing WLCG infrastructure for HEP, which has been already operating for years and SRCnet which is being set up and not yet in operation at the time of writing this document. Many research infrastructures lack systematic approaches for archiving workflows, capturing software dependencies, or ensuring long-term accessibility of container images and modules. Preservation responsibilities are often decentralised, unfunded, or reliant on short-term projects. There is limited standardisation across domains, and current practices do not always support the transition to heterogeneous, accelerated computing environments or new data spaces.

Community Evidence

The SPECTRUM Technical Blueprint, particularly Section 2.5, highlights wide disparities in software stack management across e-Infrastructures; urgent community requirements for common, portable, persistent software environments; gaps in workflow preservation, container longevity, and cross-architecture compatibility; inconsistent support for portable build systems and container runtimes; lack of long-term archival infrastructure for software artefacts; and risks associated with replacing or decommissioning HPC systems without preservation strategies. This has been highlighted by ESFRI related Research Infrastructures and there is an expectation to get those needs supported by compute and data infrastructure providers. On the user side the current legacy know-how spanning from the HEP community in collaborating from many years is providing pathways and learning for the set up of the systems for RA.

Technical Foundation

This priority is captured in the following SPECTRUM Compute and Data Continuum building blocks (D6.1 Section 4): Software & Execution; Compute Continuum.

European Alignment

Software presentation is part of the policy objectives for promoting FAIRness of scientific outputs being pursued under EOSC. EuroHPC needs to ensure portability and continuity across sites and successive upgrades of the supercomputing facilities. ESFRI priorities also focus on long-term sustainability, reproducibility, and cross-RI interoperability. This priority is strongly aligned and supported by the work done under the EVERSE project which can bring solutions and recommendations towards faster achievement of priority outcomes. Advocacy groups such as DORA or ADORE from Research Software Alliance or ESCAPE ESFRI Cluster have also been actively promoting solutions on SW sustainability

Target Outcomes and Impact Assessment

Strategic Outcomes

A sustainable, Europe-wide ecosystem in which scientific software, workflows, execution environments, key enabling SW technologies and containerised stacks are archived, portable, and reusable across generations of HPC/HTC/Cloud systems.

Success Indicators

Qualitative measures showing progress toward outcomes

• Major HPC/HTC/Cloud centres adopt harmonised software preservation policies, including documentation as integral part for of preservation policies / SW management plan
• Community-standard container images and software stacks remain executable across hardware generations.
• Increased reuse of preserved software and workflows across projects and infrastructures.
• Reduced effort required to port legacy codes to new architectures.

European Impact

Software preservation enhances Europe’s scientific autonomy, reduces risk during technology refresh cycles, and lowers long-term operational costs.

Implementation Pathway and Stakeholder Roles

Implementation Phases

Short-term (1-3 years)

• Development and consolidation of guidelines, best practices and policies for research software development that include software preservation to be included as integral part part of SW management plans and related documentation for organizations and initiatives
• Promotion of open source/open standards policies to prevent sovereignty issues and avoid proprietary vendor lock-in
• Establishment and provision of execution environments offering harmonising software stacks across European infrastructures.
• Deploying knowledge hub - in which know-how of different clusters is being placed including harmonized training programs on research software best practices across communities

Medium-term (3-5 years)

• Consolidation of standardized community-based container images and software stacks that can remain executable across next generations of heterogeneous hardware systems.
• Develop plans for software sustainability as an integral part of the preparation for scientific research programmes (including simulations, data analysis) and of the preparation of major experiments in particular
• Training programs on research software best practices integrated into european university curricula

long-term actions (5+ years)

• Ensure the provision of long-term funding is kept for software preservation across scientific disciplines
• Career paths for research software engineers clearly defined, recognised under standard tools like ORCID.

Stakeholder Roles

Stakeholder group responsibilities and benefits

Scientific Communities

Responsibilities: Define domain-specific preservation needs, share best practices, adopt common containers and workflow formats.

Benefits: Long-term stability, reduced porting effort, reproducible science, stronger global collaboration.

Research Infrastructures (ESFRI RIs)

Responsibilities: Provide domain repositories, guarantee metadata quality, integrate software and workflows into preservation systems.

Benefits: Improved sustainability, FAIR compliance, interoperability across Europe.

Computing & Data Service Providers / e-Infrastructures

Responsibilities: Maintain harmonised software stacks, build and preserve container environments, offer long-term archival services.

Benefits: Reduced user support burden, stable environments, easier migration to new HW.

Policy Makers & Funding Bodies

Responsibilities: Provide long-term funding, define sustainability mandates, ensure cross-programme alignment (EOSC, EuroHPC, National).

Benefits: Greater return on investment by fostering reusability of scientific software. Increasing European competitiveness,

Resource Needs

Investment dimensions, intensity levels, and funding alignment

• Heterogeneous compute infrastructure with fully updated software stacks able to provide proper execution environments to keep research software operational over generations.
• Manpower including infrastructure operators, software developers to operate and maintain software stack and to develop and maintain the research software

Critical Dependencies

Key prerequisites and coordination requirements

• Cross-priority dependencies:
• P04- Heterogeneous computing integration,
• P08-Software Portability,
• P09-Open Science and Reproducibility,
• P13 Skills, Careers and Training