Page tree
Skip to end of metadata
Go to start of metadata
Title (goal)Support for heterogeneous file size and quantitiy
Primary ActorResearcher
ScopeSystem black-box

Current research shows that approximately 80% of research data is less than 100MB is size, however with increased production of born digital research content combined with data from sciences, we are seeing a wide variety of data files exceeding multiple gigabytes and in a variety of formats. Researchers need to be able to submit both files by both manual upload and through the use of a bulk upload mechanism and expect reasonable performance.


  1. 1

    The ability to handle research data give us a lot of complex problems at Tufts. For example, what's a reasonable way to manage a live dataset with help from Fedora? how about a partial set? A set of research data which is partially embargoed?


    our original goal – treating research data as clean packages of data from completed research, of a similar type, with similar rights – turns out not to reflect the reality of researchers, who often work on longitudinal studies or whose buckets of "research data" might include videos with or without releases,information obtained from other sources, etc. Much of this needs to be dealt with at the appraisal/acceptance end, but if we can't do something with live data sets we're not of much use to the researchers.

  2. reasonable performance

    In terms of requirements for the Fedora 4 implementation, can you offer more detail on what would be considered "reasonable", Robin Lindley Ruggaber?