Storing and backing up data

Storing data during a research project has to be a flexible process. Data should be easy to reach, but at the same time safe and robust. Broadly speaking, the choice can be made between local, network and cloud storage.

In addition, you should also think about an appropriate backup strategy.

Local storage

Local storage refers to storing data on a physical device you have direct access to (i.e. your PC, laptop, external drive, USB-stick). Local storage is the standard way of working for many researchers. However, because storing data on local devices can be risky in terms of data loss and privacy protection, the use of network storage deserves serious consideration.

Network storage

Network storage refers to storing your data on a Ghent University server connected to your local computer over the university network. Typically, these are called ‘shares’. Ghent University offers shared network drives to its staff especially for project related storage of data.

A big advantage of using shares is the transfer of responsibility. This means that Ghent University's ICT Department (DICT) is in charge of the security, maintenance and the back-up of your data on the share.

DICT offers two types of network shares, a basic version (also called bulk or project shares) and an advanced version (ACL-shares).

The basic version will work for most research projects. While its initial capacity is limited, quota can be increased upon simple request. It is also a safe option for confidential data because the data stored there do not leave the walls of the university.

Disadvantages of using the shares are that collaboration is only possible within Ghent University and that you have to be online to work with the data.

Two types of network shares are provided by DICT:

  • Bulk shares: intended for projects needing basic sharing capacities (e.g. for a PhD-student wanting to share data with promotor). Bulk shares can easily be requested online, up to a capacity of 2TB. For larger capacity needs, contact DICT.
  • ACL shares: intended for research groups or complex projects needing more fine-grained control over read/write permissions to files and folders. This type of share requires more specialist knowledge for configuration and maintenance. To request ACL shares, contact DICT.

Before requesting network shares discuss within your research group which strategy to use, e.g.:

  • Each PhD student requests a basic share shared with the promoter. After finalizing the PhD, all data are transferred to a share of the promoter.
  • An ACL share is set up within the research group and folders are created for each person with well-defined permissions. Someone has to be appointed to configure and maintain the access control list.

How do you use network shares? Two workflow scenarios are recommended:

  • Work on share, keep no local copy.

This means that all work is done directly on the network share and no data is stored on your local computer. The advantage is that security and backups are taken care of by DICT. Also in terms of security this is good, since there is no local copy.

  • Work locally, sync files with share.

Store data on your local device (e.g. desktop computer) and use the network share for syncing. In comparison with the previous option this has the advantage that the data are available off-line. However, keep in mind that syncing your local folder with the network share needs to be configured. Also, since a local copy remains on your hard drive, this poses risks for confidential data.
Note: Syncing is not the same as making a backup. You should not make a backup from your local data to the network share as the network share is already backed up.

In some situations, storing research data on a network share for daily work is not an option. For instance, if you are processing very large data sets for analysis (e.g. fMRI data or video material), working online would be to slow to be practicable. In such situations, storing data locally is inevitable but some best practices can be kept in mind:

  • Do not forget to back up. A possibility is to sync your local copy with a copy on a Ghent University network share from time to time.
  • If you work with confidential data, take the privacy of your research participants serious. If possible encrypt the data.
  • For advice on complex setups, contact or talk to your local IT staff.

Cloud storage

Cloud storage refers to online services like OneDrive for Business, Dropbox, Google Drive, and the likes. Storing (research) data in the cloud is very convenient (accessible anywhere, syncs automatically with local storage, facilitating off-line use).

At Ghent University, researchers are only allowed to use the OneDrive for Business cloud service to store and share research files. When storing confidential or sensitive data on this service, it is advised to encrypt the data first.

More information is available on the DICT Helpdesk page about OneDrive for Business.

Sharepoint 2.0

Sharepoint 2.0 is a platform allowing easy information exchange. Although it is mainly set up with the MS Office ecosystem in the back of the mind, research data files can also be stored on the system.  The Ghent University sharepoint 2.0 instance is set up ‘on premise’, meaning that the data are stored within Ghent University.

One advantage of this system is that external users can be given access too. Therefore, sharepoint 2.0 might be the right option for you when scientific collaboration with external persons is essential.

More information is available on the DICT Helpdesk page about Sharepoint.

Backup strategy

If possible, work and store your research data directly on a DICT server (e.g. network share, sharepoint 2.0). The responsibility for making back-ups is then taken care of by DICT. On the shares, ‘snapshots’ of the data are made. Snapshots allow you to retrieve (older versions of) files, without having to contact DICT. See the instructions on how to do that.

However, sometimes storing a local copy of your data on your computer is inevitable. In that case, making regular backups of your research data is essential.

This is what you can do.

The network shares provided by DICT are not intended for systematic, incremental back-ups. This because the shares themselves back-up successive copies of the files you store there.

Instead of making back-ups to network shares, it might be a good idea to make a synchronized copy of your local data to a network share. Backup-up of these files is then taken care of by DICT.

Different tools exist to synchronize files.

or an additional layer of security, it is a good idea to make regular incremental backups to an external device as well (e.g. external HD). Best results for doing this can be obtained by the backup software from the operating system you are using.

Important: Once you have set up the back-up or synchronizing of your research data you should put it to the test. It is important to check if and how you can retrieve files form your back-up.

(Big) File transfer

Sending (big) files to collaborators can be a hassle. Especially when the files you want to share are not on a shared platform (e.g. network share, onedrive for business).

Filesender is the tool if you want to send or receive (big) files. It is secure (additional encryption can be used) and flexible as it allows you to share files with external collaborators.