In early summer 2023, the Rockefeller Archive Center was finalizing the development of its new cloud-based infrastructure for digitized audiovisual materials. This corresponded with the beginning of my 10-week Fellowship at the RAC. As a recent graduate of The L. Jeffrey Selznick School of Film Preservation, I arrived here with prior experience (primarily) in the physical and intellectual control of analog audiovisual collections. My RAC Fellowship has allowed me to continue building these skills with various film inspection and vault management projects. However, the Fellowship has also allowed me to broaden my experience with digital asset management as an integral player in the testing of the RAC’s new digital infrastructure. This has been an opportunity to develop a working knowledge of command line tools and to become familiar with digital preservation workflows.
The Rockefeller Archive Center has a long relationship with Archivematica; I think we were one
of the first institutions to implement Archivematica in production. Our use of Archivematica
coincides with our use of PREMIS Rights. Early on, we decided that we would always add PREMIS
Rights statements to transfers to support mediated access to digitized and born-digital records.
Having machine-actionable rights statements allows us to flexibly develop and change our
implementation of mediated access without being tied to a particular model or definition.
As part of scaling up our digitization and born-digital transfer processes, we wanted to distribute
the work of authoring and managing PREMIS rights statements across the organization. However,
particularly when serialized in XML, a significant degree of specialized expertise to understand
and author PREMIS Rights. Over the years we’ve created several different kinds of graphical
user interfaces aimed at lowering the barrier to the human management of PREMIS Rights.
I’ve now worked at the RAC for eleven years, which is long enough that several processes that I helped to define
and put into place need to be re-examined and completely overhauled. This has been a thought-provoking (and to be
honest, feelings-provoking) process, and has got me thinking a lot about what it means to acknowledge and deal
with things that aren’t working.
This is not a blog post about our successful implementation of Kubernetes, but rather how and why we decided not to migrate to Kubernetes after two years of work. I will attempt to call out our hypotheses and acknowledge where our thinking was wrong, and ultimately talk about the difficulties we faced and lessons we learned along the way.
The Rockefeller Archive Center is developing a system wherein audiovisual file delivery, validation, quality control, and affixing rights statements can occur entirely in a Cloud setting.