CHIN Digital Preservation Case Study – 8th Hussars Museum
Table of Contents
CHIN would like to thank Major Tom McLaughlan and all volunteers of the 8th Hussars Museum and Archives of Sussex New Brunswick for their willingness to pioneer digital preservation in small Canadian Museums, and for their offer to share the details of their efforts so that all may benefit. CHIN is also grateful to William Vinh Doyle of the Provincial Archives New Brunswick (PANB) for his invaluable contributions, as well as the Council Archives New Brunswick (CANB) for their joint involvement with the 8th Hussars museum. Finally, CHIN would like to thank members of the National Capital Region’s Digitisation Discussion Group (DDG)Footnote 1, and Canada’s Digital Preservation Discussion Group (DPDG)Footnote 2 for their contributions.
Introduction and Project Background
In 2011, CHIN conducted a survey on the state of digital preservation in Canadian Museums, and found that while museums often held digital assets, almost no museum had a formal policy or plan for the long-term preservation of these assets. In response, CHIN developed a Digital Preservation Toolkit, consisting of resources produced by both CHIN and its partners, which helped museums create digital preservation policies, plans, and procedures.
However, the toolkit did not provide concrete examples of what these policies, plans or procedures should look like. While such finished products will be different for each museum, CHIN understands that one of the best ways to learn is by example. CHIN also wanted to learn how to improve the toolkit, and so invited museums to take these tools up, and provide feedback. The 8th Hussars Museum in Sussex New Brunswick is the first known example of a small museum applying the resources in CHIN’s Digital Preservation Toolkit, and this case study is an account of that experience.
Background of the 8th Hussars Museum
The 8th Hussars Museum, which records and celebrates the 167-year history of the 8th Hussars regiment of the Canadian Armed forces, is representative of many of Canada’s smaller museums. It is located in a donated space (in this case, a preserved train station in downtown Sussex, New Brunswick), and is run entirely by volunteers. It is open seasonally to the public, and it relies on grants for the employment of students during summer months.
This museum also manages an archive and has recently digitised over 7600 images from its collections, and this has been a motiving factor in their decision to develop a digital preservation policy and plan. The museum works with the Council of Archives New Brunswick (CANB) to manage its archival holdings, and is currently storing its records on a FileMaker Pro database developed for the New Brunswick Girl Guides, a CANB member; CANB manages an online database, and encourages all its members to upload their archival content to it. The 8th Hussars also uploads some of its museum collections content (which is managed on a copy of the Virtual Collections CGI software package) to the Artefacts Canada database.
Discussion of Project Activities
During the initial development of the plan, CHIN attempted to identify a solution that would involve the OAIS standard that is commonly observed by the archival community. Artefactual, a software development firm out of Vancouver, was invited to present their open sourced Archivematica software at the Digitisation Discussion Group (DDG) and Digital Preservation Discussion Group (DPDG) joint meeting, and it became quickly apparent that this system was too complex for a smaller museum to manage on its own. Even for larger museums, some development work would have to be undertaken to make the archived material searchable, as data formats for collections management systems were not currently recognisable by this system. Comparable archival systems presented similar problems, and it was clear that a simpler solution would be required for small museums.
During CHIN’s October 2014 meeting with 8th Hussars Museum volunteers, it was suggested that maybe the best solution was “to do nothing”, meaning that the best solution may be one in which risk is mitigated as much as possible, by changing current practices as little as possible. This became the goal of the project.
Firstly, risk needed to be assessed, and doing this by taking stock of the museum’s digital assets was the first step of many that are laid out in CHIN’s Digital Preservation Toolkit.
Using the Digital Preservation Toolkit
The basic steps to digital preservation include:
- Taking stock of a museum’s digital inventory;
- Producing a policy (i.e. what you want to do, and why);
- Producing a plan (how you will do something); and
- Developing procedures (the established protocols for carrying out a plan).
To take stock of a museum’s inventory, CHIN and the volunteers at 8th Hussars Museum used the Digital Preservation Inventory Template to consider the risk and impact of losing access to digital resources. If the volunteers decided there was sufficient risk or impact to proceed, Nancy McGovern’s Digital Preservation Policy Framework would next be used to draft a digital preservation policy. In turn, CHIN’s Digital Preservation Plan Framework would then be used to produce a plan, so that the policy could be carried out. Other documents were drawn into this process, including further resources in the toolkit, as well as some from the museum itself. The following is a summarization of the Digital Preservation Toolkit process.
Step 1 – Taking Stock of the Digital Inventory Held by the 8th Hussars Museum
The printable version of the DP Toolkit’s Inventory template meets web page requirements. However, it became quickly evident during the October 2014 meeting with 8th Hussars volunteers, that an Excel version of this template would be more useable. Fortunately, both CHIN and the Canadian Museum of History had been working on such a version, and this will be made available online. In the meantime, a copy of this template can be obtained by sending a request to pch.RCIP-CHIN.email@example.com.
|Name of Digital Asset Group||Brief Description of Group||Approximate Number of Digital Assets in the Group||Approximate Amount of File Space Required to Store Group||Minimum Number of Copies of Assets in this Group (if multiple copies are kept)|
|Group 1 - Digitized Images||TIFFs and accompanying descriptions in MS-Word||7665||179 GB||2: (hard drive on office machine & external Western Digital Hard Drive) - all onsite|
|Group 2 - Administrative Docs||Fits on three CDs||2: (hard drive on office machine & separate external admin HD) - all onsite.|
|Group 3 - Virtual Collections CGI||Filemaker Pro records of museum objects.||Fits on one CD||3: (hard drive on office machine, CD with provincial heritage branch, & uploads to Artefacts Canada).|
|Group 4 - Archive Database - Girl Guide/Provincial Archives Application.||Filemaker Pro records of archival materials: papers, photos, maps.||Fits on one CD||3: Hard drive on archival machine & separate external, CD onsite & CD with provincial archives.|
|Group 5 - Interviews on CD||Only Copies of interviews regarding history of 8th Hussars||Fits on five CDs||One copy only, onsite. Some on external hard drive.|
|Group 6 - Images on CD of Oversized Photos||Only copies of oversized photos.||Fits on three CDs||One copy only, onsite.|
|Group 7 - Film converted to video on CD||Original film is now with provincial archives.||Fits on four CDs||One copy only, onsite.|
|Group 8 - Non-digital assets: 35 mm slides, 16 mm 8 mm Super 8 film, Newsreel cassettes, Audio cassette tape.||Plans being made to digitize 35. (Costco). These assets but will affect any long-term Digital Preservation Plan once they are digitized.||None - plans to digitize.||N/A||N/A|
As can be seen from the table, Group 1 (digitized images) is the largest category, and it is what inspired 8th Hussars to develop a digital preservation policy and plan.
During a presentation of this work to the Digital Preservation Discussion Group, it was noted that Group 2 (administrative documents) would generally not be part of the preservation procedures that are designed for cultural digital assets. However, CHIN included administrative documents in the recommended preservation plan and policy because smaller museums are unlikely to have a separate plan for the management of administrative materials, and because (as will be seen) there is little additional effort required to preserve these.
While carrying out this inventory activity, the following major issues were identified:
- CDs were not archival quality, and were sometimes the only copy of a resource;
- Where backups were made, they were not preservation copies;
- The 3-2-1Footnote 3 and LOCKSSFootnote 4 rules needed to be implemented;
- Refreshing media was not part of the existing process;
- Data migration was not part of the existing process.; and
- A regular backup procedure needed to be put into place, complete with preservation metadata.
Common Digital Preservation issues that were not found to be of concern included:
- Many file formats appeared to be suitable for digital preservation;
- Security was reasonable for the material being managed;
- Storage environment was reasonable (although an offsite solution was required); and
- There was no formal access policy, but for such a small institution, this would not appear to be an issue.
Step 2 – Drafting of a Digital Preservation Policy
CHIN used the framework provided by Nancy McGovern as a basis for a draft document, and it quickly became apparent that this framework was too detailed for a museum of this size. Nevertheless, the level of information provided piece of mind that no factor would be left unconsidered. All components of the framework were reviewed by CHIN, but only those relevant to the museum were used. Those of greatest interest were:
Scope: The proposed scope of the policy was to provide access to all digital assets except administrative materials for an indefinite period. Group 2: administrative materials, would be kept for 7 years, unless there was important business value tied to a given document.
Operating Principles: It was proposed that whatever procedures are put in place, they must not consume more time or resources than are already being used. While this makes a solution that fully observes OAIS (Open Archival Information Systems) model unlikely, several commonly observed digital preservation principles were retained. Namely:
- Digital content should be created with preservation in mind (eg: maintain some metadata relevant to preservation).
- Preservation copies of material should be made on a regular basis.
- An automated process should be used to create basic preservation metadata as preservation copies are made.
- Preservation copies should be kept on multiple forms of physical media, in multiple locations.
- Preservation copies should be periodically verified for data integrity.
- Physical media holding preservation copies should be refreshed to new media on a regular basis,
- Preservation copies should be migrated to new formats as required.
- Access to preservation copies should be limited to specific staff.
Selection & AcquisitionFootnote 5: It was proposed that if a digital asset is unique and falls into any of the groups in the digital inventory (other than group 2) then it should be preserved indefinitely. Items in group 2 with business value should be preserved a minimum of 7 years. Anything else may also be preserved if it reduces effort (i.e. a “batch” process).
Challenges & Risks: The greatest challenge identified was regarding limited resources. Secondly, there was a need for training, and CHIN has offered to help with this.
In terms of risk, the greatest concern was regarding content that resides exclusively on read/write CDs. These needed to be moved to a more permanent physical carrier. There was also a risk of the unknown: currently no standard exists for digital preservation in small museums, and thus whatever solution was chosen, it would involve a degree of uncertainty.
Financial Sustainability: This was not deemed to be a risk, as the proposed solution required little or no additional financing.
Technological & Procedural Stability: The template also called for an assessment of technological and procedural stability: i.e. how easy is it to maintain technology and procedures in the long-term. It is expected that both the technology and procedures would be kept simple (by necessity) in any final solution, but this component of the policy was left until such procedures were actually developed and approved.
The most recent draft of this policy is with the 8th Hussars museum for their review.
Step 3 – Drafting of a Digital Preservation Plan
CHIN drafted a plan document for the 8th Hussars museum based on the Digital Preservation Plan Framework found in the toolkit. The resulting Digital Preservation Plan is currently with the 8th Hussars museum for approval.
Five solutions were considered, and a summary of these is presented here:
|Option 1||Option 2||Option 3||Option 4||Option 5|
|Multiple Backups.||Multiple Backups.||Multiple Backups.||Simple, but fully compliant OAIS model managed internally.||OAIS model managed Externally.|
|Offsite archive. Preservation Copies for other content.||Offsite CANB archive. Preservation Copies for other content.||Offsite CANB archive. Preservation Copies for other content.|
|Checksum Generator.||Checksum Generator.|
CHIN recommended Option 2. Options 3, 4 and 5 were ruled out given that:
- There is little advantage to the introduction of a quarantine machine into the museum’s current workflow;
- It is not realistic to expect a smaller museum to run its own OAIS-compliant archive; and
- There is currently no externally run OAIS-compliant archive that is both willing and able to take on the burden of guaranteeing long-term access to the museum’s digital assets for free.
Both Options 1 and 2 require little effort beyond what is already expended by museum volunteers, and both options improve immediate search-ability of high resolution images by indexing these in the archival database so that the database can be used as a search tool to retrieve these images.
Both Options 1 and 2 also improve the likelihood of long-term access by proposing multiple backups and offsite storage. However, without significantly more effort, Option 2 offers fixity checking; a best practice that helps verify that files have not changed since they were first preserved.
An interesting item in the details of option two was CHIN’s initial recommendation to use archival (gold) CD for one set of the preservation copies; this physical media is an archival standard. However, upon presenting the proposed solution to the Digitisation and Digital Preservation Discussion Groups, members questioned the wisdom of using a physical carrier that may soon be obsolete. The argument among digital archivists is that CDs in general are being replaced by cloud storage, and the hardware to access the CDs may soon go the way of the floppy drive. On the other hand there is a relevant difference between the two technologies; floppy disks were primarily used to store software, which was vulnerable to Moore’s “law”. As both hardware and software changed to keep pace with this law, the contents of floppy disks became obsolete. There was therefore no reason to read older disks once this physical carrier had been replaced. CDs however contain a form of content (music, movies, etc…) which remains valuable for longer periods of time, in spite of newer standards or more convenient ways of storing this content.
This debate remains unresolved, but for the sake of prudence and (as it turned out) simplicity, CHIN changed its recommendation to the use of external hard drives in lieu of archival CD. This contravenes the 3-2-1 Rule (i.e. keeping preservation copies on at least two physical formats), but it was considered the lower risk solution, and one that was easier to implement. Details of this recommendation can be found in the 8th Hussars Digital Preservation Plan.
Developing General Procedures for the Recommended Plan
Once a general plan was in place, procedures to carry out the actual work needed to be established, and the advantage of choosing external hard drives over archival CD became apparent; external hard drives are far simpler to manage than a multitude of CDs.
Ingesting content piecemeal (as is typically done in a digital archive) was not tenable, as the work required would soon outstrip the volunteer labour available. Likewise, making a single preservation copy of all content at the outset would not work, as the content was constantly being edited, and added to. Making multiple preservation copies would soon fill hard drive space (even given the exceptionally large 2 TB disks that were available at the time). And so, a hybridized solution was considered, using personalised backup software.
This software would be used to create an initial preservation copy of all digital asset groups, then add to it incrementally, as new files in these groups were created. Databases (which consumed less space than images) would be copied manually on a periodic basis, and multiple copies of these could be retained. The details of the proposed procedures can be found in the Digital Preservation Plan.
One interesting note during the selection of personal backup software was that vendor claims did not always match CHIN’s own observations. For instance, the ability of the software to distinguish between the creation of a new file since a previous backup, and the renaming of an existing file could not be reproduced on a majority of packages that claimed this feature. This would be important (for instance) if image files were renamed – backup software that did not have this feature would keep two copies of files in this asset group.
Another interesting note was the vendor claim of using MD5 checksums (an algorithm often used to assess file integrity); while packages did appear to use a data integrity checker, they often stored this information in a proprietary format – making it difficult to access the checksum information without use of that particular software package. This is fine for short-term backups, but in the long-term, it would add the otherwise unnecessary step of migrating data, should that particular backup software no longer be supported. To resolve this, a second checksum generator was used which stored the resulting checksum information in an easily accessed standard text document.
Current Status of the Project
CHIN has presented its final recommendations to the 8th Hussars museum, and to the Digitisation and Digital Preservation Discussion Groups for their final thoughts. CHIN will then assist the museum, should they decide to implement the recommendations. A follow-up may also be carried out within a few years’ time. Since having submitted its recommendations, CHIN has learned that 8th Hussars will be using CANB’s online database exclusively for management of its archival holdings (i.e. the museum will no longer maintain a second computer with a local database for digital asset group 4). CANB will be fully responsible for preservation of this data. This greatly simplifies CHIN’s recommendation, but the Plan and Policy documents published on CHIN’s site will include a recommendation for how Group 4 assets might have been managed by the museum.
Next steps involve CHIN assisting the museum, in implementing these recommendations (or whatever variation is most suitable for the museum), and helping museum volunteers with software training.
Contact information for this web page
This resource was published by the Canadian Heritage Information Network (CHIN). For comments or questions regarding this content, please contact CHIN directly. To find other online resources for museum professionals, visit the CHIN homepage or the Museology and conservation topic page on Canada.ca.
- Date modified: