Skip to content

Backup

Use this page to decide how to protect files against accidental deletion, device failure, theft, and ransomware.

Core principle

Do not assume a file is protected just because it exists somewhere in the University or on a managed device.

In practice, backup planning depends on where the data lives:

  • data on a local desktop or laptop disk may not be backed up at all
  • cloud synchronisation is not the same thing as a full backup strategy
  • high-performance research storage may prioritise speed over backup
  • archive storage is not always suitable for day-to-day recovery

Local computers

Any data stored directly on a desktop or laptop should be treated as at risk unless you know a backup service is in place.

Risks include:

  • hardware failure
  • theft or loss
  • ransomware or malware
  • accidental deletion
  • operating system rebuilds or device replacement

If the data matters, keep the master copy on an appropriate central storage service or make sure a backup product has been configured for that machine.

OneDrive and Teams

University cloud services can help with resilience, but they are not a complete substitute for backup planning.

OneDrive

OneDrive provides version history and short-term recovery options, which can be very useful for accidental changes or deletion.

However:

  • files that are cloud-only may not be included in local device backups
  • synced folders are still vulnerable to user error and some malware scenarios
  • OneDrive data is tied to the user's University account lifecycle

For particularly important files, make sure they are also stored in an appropriate managed location or copied to another protected store.

Teams

Teams storage also includes recovery and versioning features, but it is collaborative storage rather than a dedicated backup system.

It is useful for shared working documents, but important research data should still have a clear backup or archive plan.

Research storage

Different research storage platforms have different protection models.

High-performance working storage

Fast research storage, including cluster-attached storage, may not provide backups. Treat these platforms as working storage unless documentation for that service says otherwise.

They are often suitable for:

  • active analysis
  • temporary processing outputs
  • staging data for compute jobs

They are often not sufficient on their own for:

  • irreplaceable derived data
  • sole copies of research outputs
  • long-term retention

Archive storage

Archive systems are useful for recovery and retention, but they are not always designed for frequent day-to-day access.

Use archive when:

  • you need a retained copy of a completed dataset
  • you want an additional copy separate from active storage
  • the data should be preserved beyond the active analysis phase

For OxCIN-hosted tape archive workflows, see Archive.

Backup options to consider

Choose the option that matches the type of device and the value of the data.

Central file services

Where possible, keep important data on managed central storage rather than on individual computers.

Device backup software

For local computers, OxCIN guidance refers to services such as:

  • Code42 CrashPlan for some single-user computer scenarios
  • IBM Spectrum Protect / TSM for some multi-user or managed backup scenarios
  • Time Machine on macOS for local backup workflows

The exact backup product depends on the device and support arrangement, so check with OxCIN IT if you need a backup configured.

Additional encrypted copies

For especially important material, an additional encrypted copy on a separate managed storage location or encrypted drive may be appropriate.

If the data is sensitive, the backup copy must be protected to the same standard as the original. For encryption guidance, see Encrypting data.

Virtual machines

If you use a virtual machine on your computer, backup software often excludes the VM disk files.

In many cases it is better to back up the important files inside the virtual machine rather than relying on backup of the whole VM image.

Practical guidance

Use this quick checklist:

flowchart TD
  A["Where is the only copy of the data?"] --> B{"Local computer only?"}
  B -->|Yes| C["Move it to managed storage or arrange backup"]
  B -->|No| D{"Working storage without backup?"}
  D -->|Yes| E["Create a second protected copy or archive copy"]
  D -->|No| F{"Need long-term retention?"}
  F -->|Yes| G["Use archive or another retention service"]
  F -->|No| H["Review versioning and recovery options"]

Good practice

  • Keep scripts and code in version control, for example GitLab.
  • Keep irreplaceable derived outputs somewhere with a clear recovery path.
  • Do not rely on a single laptop, desktop, or external drive.
  • Review backup arrangements before staff leave the University or accounts are closed.
  • Be especially careful with cloud-only files and virtual machine disk images.

Need help?

If you are unsure whether a storage location is backed up, or you need backup arranged for a local computer, contact OxCIN IT.