[Rdap] Getting into data hosting

[Gulliford, Bradley]

Hello all (and please excuse the cross-posting to IASSIST-L)--

My library wants to start providing data hosting--acquiring hardware and inviting departments (student and faculty) to store their research data with us (saving them cost, managing format migration, preservation, etc.)  Up to now we have avoided that, like most libraries, and concentrated on providing navigation (finding or linking to data stored elsewhere) and general advice.  Now we want to buy (collocate or purchase cloud capacity) actual computer systems on a much larger scale than we have ever done.  (Even servers were handled mostly by our university IT department.)

If you have coordinated a major installation (or you have knowledge of how it's done), could I trouble you for some suggestions?  Specifically my questions at this time are:

1.  How did you estimate capacity required (number of bytes)?  Did you just start with a number and revise it with experience, or did you do a comprehensive data census, or did you ask a few departments what they had on their PCs?

2.  Did you make any arrangements for access (availability or control) other than logging regularly into your servers/hosts?

3.  What kind of relationship do you have with your university IT department?  Do they provide services you use?

Obviously we have many more issues to consider (organizational politics, retention and versioning policies, and preservation among others) but right now we're looking at capital assets.  You can reply to me and I will summarize for the list.  Thank you.

Brad Gulliford
Data Management and Curation
University of Texas at Arlington Library
Arlington, Texas, USA
817 272-7156