Not Your Childhood Fire Drills

Bradley Daigle is the Content Lead for the Academic Preservation Trust and Chair of the Coordinating Committee for the NDSA's Leadership Group

While at the most recent PASIG, held in Mexico City, I sat in the audience and responded to a question that I interpreted to be about preservation stewardship responsibility and where it resides. Apparently, the solution I put forward was not something that others had considered or were planning. Therefore, in this post, I am providing some context and explanation of what we are doing at Academic Preservation Trust (APTrust) to provide a more rounded approach to preservation responsibility. This overview will be a description of what I call “fire drills” or what our tech team refers to as “test restores”.

What are we talking about?

When it comes to taking responsibility for preservation activity, the duty can be assigned in many ways. It can be assigned to the individual within an organization, the organization as a whole, or the preservation service if it is an external entity.  My experience over the last few decades has shown that responsibility for what preservation actually means can frequently shift depending on one’s organization and perspective—and this shift can often be imperceptible, going unnoticed until a crisis occurs. I don’t just mean the “its not my fault’ approach to whether or not preservation has been successfully executed but rather the why and how we are preserving our cultural heritage as a profession. Preservation is an active set of actions rather than being passive. Active in the sense that it is an ongoing activity - one that never truly is completed, an asymptote in its purest form. Not passive in that preservation is not just storage. The act of putting collections into storage should not be construed as true preservation without the active management of those materials (fixity, versioning, etc.) and that management assigned to various staff within an organization.

Organizational responsibility

There are some clear indicators that comprise a comprehensive organizational approach to preservation. There is no shortage of primers on the subject so I will not go into detail in this post but a simple checklist can look like this:

1. Collection Development/Curatorial Policy [what do you collect?]
2. Preservation Policy [how does your organization define preservation?]
3. Technical pathway for preservation (think the Levels of Preservation) [how is a preservation strategy applied?]
4. Clearly assigned roles for updating / managing each of these components [who does what and with what kind of accountability?]
5. Review schedule [how often do you need to revise this process?]

By no means a comprehensive list, these steps are a good place to start for any organization. There is another obvious element that runs through all of these components, and that is transparency. Without the means to clearly verify what is being expected and undertaken then accountability disappears. In particular, when an organization relies on a third party service, that transparency and accountability become a critical component. 

Responsibility of the service

There are some basic elements that are hallmarks of a successful preservation service. These can range from the highly detailed compliance with ISO 16363 to a generalized solution for a given organization. Many of the basic principles noted above still apply. Steps 1-2 take place locally but perhaps the third step is undertaken by an external service. That would make numbers 4 and 5 shared between organization and service. Clarity among all the roles and specific responsibility for each action is perhaps more important when it is undertaken by an external service. If that service is either unable or unwilling to provide specific details on what that means for what you need, that should be seen as a warning sign - and a violation of that critical component, transparency. 

With Great Responsibility…

Next I want to talk about transparency versus assurance. In some cases, vendors will not provide the level of transparency commensurate to their preservation responsibility. Or if they do, it is not widely shareable due to some form of non-disclosure agreement. This can be for reasons that are quite sensible - some of these solutions fall under the rubric of trade secrets, the sharing of which would put that service’s existence at risk. As understandable as this reality is, how do we, as a profession, continue to absorb responsibility for preservation if we cannot know how it is being done or cannot share our stories about its failures? This is where assurance comes in. This call to my preservation colleagues was made at PASIG as well. While I understand the need to obfuscate certain technical details within preservation solutions we, as a profession dedicated to preservation, need to understand more fully how to work with services that limit the sharing of necessary information. We need to develop assurance metrics - metrics that can be clear and verifiable but not wholly revelatory of trade secrets. It is up to us to do this because the private sector will not. Otherwise, we need to think long and hard about using services that do not allow for open sharing of how preservation takes place.

Active Preservation - APTrust “Fire Drills” as Content Restoration Tests

In my opinion, one of the best tests for assurance is what I refer to as “fire drills”. These can also be referred to as restoration spot tests. Essentially, this means that from time to time a service requires a depositor to restore some of the content that they have submitted through that service. This has a trifold purpose. The first is that it provides a verifiable action that the content is what it is expected to be - thus giving the depositor greater assurance that the materials are being preserved properly. This is a critical test. A depositor should always get they expect. Anything else compromises the assurance of the service. The second purpose, of perhaps even greater value is the test these restorations provides for the organization that owns the material. In other words not only should the depositor get back what was expected but were they able to understand and identify their own materials—particularly if they were deposited by previous staff or from a different department. That understanding is as much a test of the depositor’s active preservation practices as it is of the service that holds their content. Finally, this effort is a robust test of a system’s ability to do restorations easily and reliably as part of a suite of best practices. Therefore these fire drills underpin an active preservation strategy from many sides of the service and organization.

After some initial testing, APTrust is working with its members to determine a frequency that may allow for some configuration of these restorations. For example, a new member might request a greater number of fire drills initially to test both its own workflows as well as the assurance of the service. Then over time, the frequency may decrease as confidence rises. There is also a need to balance the costs of moving materials around on both sides (staff time and I/O costs) so the right balance must be struck. Initial results have also provided some insight into how depositors bag and submit their collections. Obviously, if a collections package is several terabytes this would not necessarily be something that should be restored frequently. Therefore taking into consideration how and in what ways these restorations will test a depositor’s local preservation management will be part of the equation. As APTrust moves forward with these innovations, we encourage other services to explore the means by which we, as a community, can align around fire drills as a best practice. We will report out more as we delve deeper into this engagement with our members.

Cause for Danger: Real Fires

As if the need for preservation failsafes and transparency weren’t clear enough, the recent investigations of the “Universal Studio Fire” that happened in 2008, the extent of which was effectively suppressed, is only now becoming truly known. How many more examples like this need to be brought to light? In many ways, the preservation community still functions like the special collections and archives community of decades past - where theft and disaster lessons were not widely shared due to fears of reputation loss and donor concerns. How could the scope of such a disaster as the Universal fire have been mitigated if the community had known earlier? Pure speculation at this point. This is why we need to encourage and support conversations around preservation disasters/mishaps/snafus in a way that we can learn and grow as a community. Organizations like the DPC work to create those spaces and opportunities to do so and continued effort along those lines needs to become ongoing, normalized practice across the wider profession. 

The work that APTrust has begun with its members and the broader community of distributed digital preservation is just one of the efforts to increase assurance and transparency of practice. It would be wonderful to hear from others who may find ideas such as these “fire drills” a useful pursuit as well as other possibilities yet to come. 

If you want more specific details as to how APTrust is doing this work, you can find it here.