Data transfer

Attention: Confluence is not suitable for the storage of highly confidential data. Please ensure that any data classified as Highly Protected is stored using a more secure platform.
If you have any questions, please refer to the University's data classification guide or contact ict.askcyber@sydney.edu.au

Data transfer

Working with research compute systems will require you to move your data between your personal computer, the systems you work on, and various storage systems. There are lots of ways to move data, Globus and FileSender as the preferred ways to move your research data securely.

Recommended approaches

The University supports multiple data movement services, the right one for you will depend on your dataset and the locations you’re moving it between. There are multiple protocols and tools available for transferring data. Each has its strengths and limitations.

Method

Use case

Benefits

Limitations

Method

Use case

Benefits

Limitations

scp/sftp

small, ad-hoc file transfers

simple, widely available

fragile for large data, no automatic resume

rsync

synchronising directories

efficient for incremental updates

manual retries, complex flags, not fault-tolerant

browser (HTTP) downloads

small public datasets

easy for one-off use

not suitable for large or private datasets

cloud sync clients

desktop-focused workflows

convenient for local files

poor fit for shared storage and HPC

managed transfer services (Globus and FileSender)

research-scale workflows

reliable, scalable, auditable

requires some setup and understanding

Please be aware that we are rolling out access to Globus for early adopters in Q1 with a broader rollout in Q2 2026. If you would like to be an early adopter, please get in touch with SIH (https://sih.tools/request).

What is required for data movement?

At a minimum, moving data between systems successfully requires:

  • A source and a destination

  • A network connection between them

  • A protocol that controls how data is transferred

  • Appropriate authentications and permissions

  • A way to handle errors, interruptions, and retries

Supported services

The University of Sydney supports data movement between a range of research computing services, including:

  • Research computing platforms

  • University-managed storage services (RDS)

  • Selected cloud platforms

  • Researcher-managed systems

For most supported services, Globus and FileSender are the recommended mechanisms for data transfer.

Source and destination

All data movement involves a source (where data is coming from) and a destination (where data is moved to). In research computing these are commonly:

  • A workstation or laptop

  • A shared research storage system like RDS

  • An HPC platform

  • A cloud storage system

  • An external institution’s system

Understanding both ends of a transfer is important because access permissions can differ between systems, performance characteristics vary, and not all protocols are supported everywhere. Managed services like Globus are designed to handle transfers between heterogeneous systems without requiring users to manage these differences manually.

Network connections

Data transfers rely on network connectivity between the source and destination. In practice network connections can vary due to network reliability, firewalls and security controls, long-distance connections, and institutional network policies.

Beware that tools that depend on continuous interactive connections are more likely to fail when networks are unstable. Managed transfer services like Globus and FileSender are better suited to research workflows because they can tolerate interruptions and resume automatically.

Managed services

Globus

See the Globus pages for details on how it can be accessed and used:

Filesender

See the FileSender pages for details on how it can be accessed and used:

CLI based methods

You may prefer to use familiar command-line tools like rsync. For more information on these methods see: