Data storage:

The 3-2-1 backup rule

During the course of a research project, numerous data sets are created in different versions. In order to work effectively and collaboratively with them, it is important to define the naming and versioning of the research data in advance. In general, it makes sense to create read-only versions of the original data and only work with copies of the master file. We recommend the 3-2-1 backup rule:

• create at least 3 copies of the file
• store them on at least 2 different storage media
• store one copy decentrally (e.g. in the cloud)

Where is research data stored?

Research data should be backed up regularly. Established infrastructures of the University Computer Centre are e.g:

• network storage
• the OwnSky cloud (cloud variant of TUBAF based on Nextcloud).

Data archiving

Preferred file formats

The data format plays a decisive role in the long-term storage, collaborative work and reuse of research data. If possible, data should be saved in open file formats (e.g. .html, .jpg, .mp3, .gif). This allows this data to be read and edited by various programmes.

Long-term archiving (LTA)

The following file formats are recommended: