In the autumn, colleagues from Gloucestershire Archives have been showcasing our digital preservation work at a couple of conferences. The first the Archives and Records Association conference was about ‘Facing Forward: Post-pandemic recordkeeping – change, challenge, choice’ and the second was the international conference on digital preservation (iPres) focusing on ‘Data for all, for good, for ever: Let Digits Flourish’
For both conferences we presented papers focusing on the recent work we have been doing designing a long term digital storage solution for our born digital records and curiously enough (or perhaps not!) this very forward facing topic served to highlight the basic principles that underpin the work we do here.
So first, what are archives?
Archives are the record of everyday activities of governments, organisations, businesses and individuals. Archives may take many different forms – handwritten, typed, printed, photographic or electronic – and include audio-visual material such as video and sound recordings. As authentic and reliable records, they are preserved permanently because of their evidential and historical value.
What does Gloucestershire Archives do?
We gather archive collections and local and family history resources to ensure they are kept secure and made accessible.
We can see the foundations of the Archive are provenance (that is understanding where something has come from) and authenticity. So an Archive’s worth is that it preserves both provenance and authentic content.
Historically then we expect that the documents we offer to customers are precisely the documents that crossed the Archive’s threshold and are what we have ever since kept safe. Archival authenticity does not mean that a document’s content is “true” rather it means that the document produced is the document that was received by the Archive, possibly many decades earlier.
Therefore successful Archival preservation requires more than just having access to a document. We must also know where the document is from and how it relates to other documents. And in particular we must know that our document is authentic and be able to prove this claim.
All then that we need to do for our digital records is translate these principles into the digital world.
We have been working on tools that will ensure that we not only know that our digital documents have been preserved but that we can prove it.
Provenance is captured in our hierarchical catalogue that is compliant with international cataloguing standards (ISAD(G)). This digital application is supported by our normal business continuity plans, and our “disorderly exit” plan – which protects us in case of incidents such as supplier failure.
Authenticity is based on the custody of the fixity digests of the archival information packages (AIPs) or digital objects that are held by us in our storage.
You can read more about fixity and fixity digests on the Digital Preservation Coalition’s website https://www.dpconline.org/handbook/technical-solutions-and-tools/fixity-and-checksums, but essentially you can think of a fixity digest as a digital fingerprint or unique value that can be generated from a digital object. Knowing, maintaining and comparing the fixity of a digital object allows us to prove that a digital object is authentic.
In practical terms we use a packaging tool to create Archival Information Packages or AIPs from the digital documents transferred to us. (Think of this a bit like putting some paper records into a box). As part of this process we calculate the fixity digests of the AIPs.
We are then able to deposit AIPs in a remote cloud based store. We use a storage fixity tool to calculate the fixity digests of the deposited AIP that it has received. It reports these digests back to the packager tool. The packager tool knows what the fixity digests of the AIP should be since it calculated them when the AIP was first created. So it can verify what the storage fixity manager tool is reporting and confirm that the service is reporting the expected fixity digests.
Finally, the packager tool includes the fixity digests that it created earlier in a fixity digest database which is maintained by the Archive.
Similarly when a user wants to consult a particular digital object we can request an AIP from the store using the packager tool.
The archivist identifies the AIP that is being requested. The packager tool receives the downloaded AIP. It confirms that it has located the expected fixity digests in the fixity digest database maintained by the Archive. It calculates the fixity digests of the deposited AIP that it has received so that these can be compared. Since the digests agree with the expected fixities that have remained in the Archive’s custody the Archive can prove that the requested AIP is authentically identical to the deposited AIP.
Gloucestershire Archives has been working in this field now for almost 20 years. Our approach of learning by doing has allowed us to develop our thinking and learning by actively taking discrete steps to preserve our digital collections. Our existing AIPs are stored securely within Gloucestershire County Council’s network, but now that we have defined our requirements we can explore the opportunities that new technologies offer.
Claire Collins, Collections Development Manager
Gloucestershire Archives 