Features | Platform Capabilities

OSP is built around four composable core modules — Apps, Compute, Data, and Publishing — supported by integrated identity management, a content management system, and a fully managed hosting model. Each module can be used on its own or combined to fit your deployment needs.

Apps — Research & Education App Store

OSP streamlines how computational tools are delivered, shared, and reused, making advanced research applications accessible through a browser without requiring command-line expertise.

• Custom App Store — Organize and publish a branded App Store featuring domain-specific tools and workflows tailored to your community.
• One-click launch — Deliver interactive web apps (Jupyter, RStudio), graphical apps (MATLAB, ANSYS), or batch executables and scripts (serial, distributed, or parallel) with a single click.
• No-code UI builder — Create intuitive web-based launch interfaces using low-code/no-code configuration. Built-in input validation ensures correctness without custom development.
• Bring-your-own-app — Deploy existing containers or scripts with minimal configuration, enabling teams to contribute and share tools without central bottlenecks.
• Reproducibility — Job parameters, software versions, and input files are tracked and stored. One-click restart and checkpoint/resume supported where the application allows.
• FAIR app publishing — Publish apps and workflows as citable digital objects with metadata and persistent identifiers.

Compute Integration — Flexible and Location-Agnostic

OSP abstracts infrastructure complexity, connecting researchers to compute resources without requiring them to manage the underlying systems.

• Universal compute connectivity — Integrate any compute cluster or standalone Linux host — on-premises, cloud, or hybrid, including ACCESS and NAIRR resources. Multiple systems per tenant supported.
• Flexible runtimes — Docker, Apptainer/Singularity containers, executable scripts, or compiled binaries. Serial, parallel (MPI), many-task, and GPU workloads supported.
• Backend-agnostic job management — Supports Slurm and unscheduled single hosts. Job submission, monitoring, and output retrieval are abstracted from the user.
• Private node access — Proxy to non-public compute nodes behind NAT; only the gateway is internet-facing.
• Account-based orchestration — Execute jobs using individual user credentials for fine-grained access control, or fixed service accounts for team-based workflows.

Tapis Integration — Job and Data Lifecycle Management

OSP integrates with the Tapis API to provide a reliable execution and provenance layer across all connected systems.

• Job lifecycle management — Monitor job status, retrieve outputs, and manage the full execution lifecycle through a consistent API layer.
• File transfer and data movement — Secure, efficient movement of data between users, compute resources, and storage backends.
• Provenance tracking — Automatically records metadata about job execution and data changes, supporting reproducibility and audit requirements.

Identity and Access Management — Secure and Federated

OSP supports diverse authentication models and federated access, from small single-institution deployments to large multi-institutional collaborations.

• Flexible authentication — Campus LDAP, SSO systems (Shibboleth, SAML, CAS), standard web accounts with optional MFA, or federated providers (InCommon, CILogon, Globus Auth).
• Federated identity — Users authenticate with home institution credentials via CILogon or Globus Auth. No shared IdP required across systems.
• SSH key bridging — Optional binding of web users to remote Unix accounts.
• Role-based access control (RBAC) — Assign permissions by role (admin, contributor, reviewer, viewer) across apps, data, and content.
• Group management — Group-based access controls are planned.

Data Sharing — Contextual and Collaborative

OSP integrates data sharing directly into research workflows, with no separate tools required.

• Granular sharing controls — Share datasets with individuals, groups, or the public with fine-grained access controls.
• Metadata and annotation — Add rich metadata, notes, and comments directly on data assets for collaborative review and refinement.
• Custom visualization plugins — Extend data interaction with tailored plugins for tabular, geospatial, genomic, or simulation data.
• Storage — POSIX and S3/object storage, and data management on multiple systems.
• Globus data transfer — Planned.

FAIR Publishing — Share, Preserve, and Cite Research Outputs

OSP provides a full publishing framework compliant with FAIR principles (Findable, Accessible, Interoperable, Reusable).

• Open access repository — Publish articles, reports, datasets, notebooks, apps, and workflows in a customizable repository.
• Persistent identifiers — Assign DOIs, Handles, or ARKs to ensure permanent citation and reliable referencing.
• Configurable metadata — Use built-in metadata standards or tailor schemas to disciplinary or institutional requirements.
• Multistage curation workflow — Draft, review, revise, and finalize publications through a configurable editorial pipeline.
• Embargo and access control — Configure embargo periods and access restrictions to align with funding policies and IP requirements.

Content Management System — Drupal-Based and Extensible

OSP is built on Drupal, providing a robust foundation for managing all site content alongside computational capabilities.

• Full content publishing — Host documentation, announcements, training materials, event listings, and community features in one place.
• Customizable design — Tailor navigation, layouts, and branding to reflect your project or institution's identity.
• Unified environment — Content, apps, data repositories, and publication workflows are integrated in a single platform rather than maintained separately.

Fully Managed Hosting

Each OSP instance is delivered as a fully managed, secure deployment — professionally hosted and maintained so your team focuses on research, not infrastructure.

• No backend management — Infrastructure, security patching, upgrades, and monitoring are handled by the OSP team.
• Modular architecture — Use only the components you need. Extend with plugins, APIs, and custom modules as requirements grow.
• Scalable from prototype to production — Supports individual researchers, small groups, departments, and national-scale collaborations on the same platform foundation.

Want to see how these features come together in a real deployment? See OSP projects → Discuss your use case

Not ready yet? See how OSP compares with other research computing platforms