We denote several types of data in the compute environment of the scientific systems:

• Science generated: New information generated by lab experiments or theoretical calculations form the basis for the scientific output of the university. This unique type of data should be stored on highly save media and kept available for at least 10 years (data management requirement).
• Science data from elsewhere: Often data is shared between institutes, where the above requirement should be fulfilled by the institute that generated the data. The institute that receives the data has no obligation wit regard to data management, but has an obligation to the users of the data.
• Science data intermediate products: While performing scientific analysis many intermediate products are created. These products can always (at certain costs) be recreated from the original data and associated software/scripts. This type of data can be denoted as scratch data for which low reliability hold.
• Science publication data: These final products are the scientifica output from the university and should be kept available for at least 10 years (data management requirement).
• Office management information: These data are personnel or personal data for which high security and high reliability are standard. The information is usually enclosed in database format. In addition software to access and manipulate this data shuld also be treated with high security and high reliability.
• System management information: This type of information is sensitive as it describes in detail the institute facilities for which many levels of security and reliability should be in place. In addition the access and manipulation tools should be highly secure.

Below a table denoting the types of data and their characteristics.