Connecting RSpace to Institutional File Stores

Overview

Researchers often have large datasets on institutional servers — gigabytes of microscopy images, sequencing output, or simulation results — that are impractical to upload into RSpace directly. Rather than copying files, RSpace can connect to your institution's existing file storage and link to files where they already live, keeping your lab notes and data provenance intact without duplicating data.

This guide covers:

• When to link vs. upload
• Setting up a file store connection (IT administrators)
• Connecting to a file store and linking files (researchers)
• Including linked files in exports
• What's coming next

When to link vs. upload

• Small files (documents, images, PDFs under ~250 MB) you want to keep inside RSpace → Upload to RSpace Gallery
• Large datasets already on an institutional server (e.g. sequencing data, imaging output) → Link via file store
• Files managed centrally by IT with their own backup and access controls → Link via file store
• Files you want to include in a portable RSpace export archive → Either, as linked files can be bundled into exports

Linking keeps your data in one authoritative location. Your RSpace documents become the research context — the notes, protocols, and metadata — while the files themselves stay on the infrastructure your institution already manages.

For IT administrators: setting up a file system

Before researchers can connect, a system administrator must configure at least one File System in RSpace.

Step 1: Enable file store support

In your RSpace deployment properties file, set: netfilestores.enabled=true. Restart RSpace for the change to take effect.

Step 2: Add a File System

1. Log in with a sysadmin account and navigate to System → Configuration → Institutional File Systems
2. Click Add new File System
3. Select a Client type — RSpace supports:
   1. SMB (Samba v1 or v2/3): Windows shared drives, NAS appliances, institutional network shares
   2. SFTP: Linux/Unix file servers with SSH access
   3. iRODS: Institutional iRODS deployments (e.g. Yoda, SURF iRODS)
   4. S3: AWS S3 or S3-compatible object stores (e.g. MinIO, LRZ Data Science Storage)
4. Fill in the Name, URL, and authentication settings for your chosen protocol (see protocol-specific notes below)
5. Set Status to enabled
6. Click Save

SMB (Samba) configuration

RSpace supports both SMBv1 (JCIFS) and SMBv2/3. Use SMBv2/3 if your storage server supports it.

• URL: the server hostname, e.g. smb://fileserver.university.ac.uk
• Domain: the Samba domain users belong to
• Share Name (SMBv2/3 only): the first path element after the hostname. For example, if your share is at smb://fileserver.university.ac.uk/research-data, enter fileserver.university.ac.uk as the URL and research-data as the Share Name.
• Authentication: username/password (users are prompted on first connection per session)

SFTP configuration

• URL: the SFTP server address
• Server public key fingerprint: required to establish a secure connection. On the RSpace server, run: ssh -oHostKeyAlgorithms='ssh-rsa' username@yoursftp.server.org Then check .ssh/known_hosts for the new entry. Copy everything after ssh-rsa (the long base64 string) and paste it into RSpace.
• Ensure the SFTP server allows challenge-response authentication (ChallengeResponseAuthentication is yes in /etc/ssh/sshd_config).

iRODS configuration

Select iRODS as the client type and enter the connection details for your iRODS (ICAT) instance. Full field reference can be found in Configuring Institutional File Systems (SMB/SFTP/iRODS/S3).

S3 configuration (RSpace 2.22 / 1.122 and later)

S3 uses server-wide authentication — credentials are set once at the server level, and all RSpace users can then access the configured bucket.

Add the following to your deployment properties: netfilestores.s3.global.credentials.accessKey=<IAM access key> netfilestores.s3.global.credentials.secretKey=<IAM secret key>

In the File System configuration:

• For AWS S3: select AWS — no URL required
• For other S3-compatible providers (e.g. MinIO, Ceph, LRZ): provide the endpoint URL and enable path-style URL access if required by your provider

For researchers: connecting and using a file store

Once your IT administrator has set up a File System, you can create a personal Filestore — a saved bookmark to a folder on that file system — and start linking files into your documents.

Step 1: Create a Filestore

1. Go to the Gallery and select Filestores from the sidebar (if you don't see this option, contact your system administrator — the File System may not be enabled yet)
2. Click Create → Add a Filestore
3. Select a File System from the list configured by your admin
4. Enter your credentials if prompted (username/password for SMB, SFTP, or iRODS; S3 uses server-wide access and requires no login)
5. Browse the folder tree and select a top-level folder as your Filestore root
6. Give your Filestore a name and click Add Filestore

You can create multiple Filestores from the same File System — for example, one pointing to your raw data folder and another to your analysis output folder.

Step 2: Link a file into a document

1. Open a document for editing
2. Click Insert → From RSpace Gallery
3. Select Filestores from the sidebar, then choose your Filestore
4. Browse to the file(s) you want to reference and click Insert

A link to the file appears in your document. Clicking the link lets you view file metadata and download the file directly through RSpace. iRODS links are marked with an iR icon; other file store links display the file name.

Step 3: Include linked files in exports

When you export a document or folder, RSpace can bundle the linked external files into the export archive — useful for sharing a complete dataset with a colleague or depositing to a repository.

In the export dialog, look for the "Include filestore links" option after selecting an export format. When enabled, RSpace fetches the files from the connected file store and includes them alongside your documents in the zip archive.

You can send the bundled export directly to repositories such as Dataverse or Zenodo. For a full walkthrough, see Bundling linked external files in your exports.

Moving files from RSpace to a file store

If you have files uploaded to the RSpace Gallery that you want to move to your iRODS instance:

1. Select one or more files in the Gallery
2. Click Actions → Move to iRODS
3. Select your iRODS Filestore destination, choose whether to retain a copy in RSpace, enter your credentials, and click Move

What's coming next

We're actively working on deeper file store integration based on feedback from institutions using RSpace alongside infrastructure like LRZ Data Science Storage, Yoda, and iRODS deployments across the Netherlands.

Upcoming capabilities we're exploring include:

• Direct upload to file stores from RSpace — push data from your ELN directly to an S3 bucket or iRODS collection, with the link automatically created in your document. This would make RSpace the active bridge between your lab notebook and institutional storage, rather than a read-only viewer.
• Metadata orchestration — automatically generate sidecar metadata files (e.g. DataCite YAML or RO-Crate JSON) alongside your data when transferring to a file store, so your files arrive at their destination with structured metadata already attached.

We'd love to hear from you. If you're using RSpace with an institutional file store, or if you have a workflow that isn't well-supported today, please reach out through the RSpace community forum or contact your account manager. Your use cases directly shape our roadmap!

Related articles

• Configuring Institutional File Systems (SMB/SFTP/iRODS/S3)
• iRODS Integration
• How does RSpace store files? Are there configurable options?
• Export Options
• AWS Gateway and linking to AWS S3 files through SMB connector (legacy setup)