Do it Yourself Repository
Define a set of local policies and procedures for file ingest, processing and storage.
- Use simple e-records survey forms and submission agreements to identify and accession records. Keep them with your other accession records (in paper or electronic form) and also put a copy inside the nascent a Submission Information Packet consisting of:
- Submission Agreement
- At a minimum, track fixity information (e.g MD5 checksums) , for the original files using one of the ingest tools, in the metadata folder ‘archival information packet.’ Put fixity info in a tab delimited file produced by something like Jacksum or Karen’s Directory Printer. This is part of your PDI (Preservation Description Information). Make sure you use the same format for each set of checksums you produce, so you can easily migrate it later.
- Use a batch program on top of FITS to generate metadata about each file. Store the resulting output it in your metadata folder.
- Decide whether or not you want to undertake preservation actions (such a migration) for particular types of files
- If you do, develop a preservation and access plan; then,
- Implement either proprietary or open source tools for preservation actions (e.g conversion/migration). Do not discard any original files! For each set of records you accession, place the originals into an ‘originals’ folders in your ‘Content” Folder. Place the converted files in a ‘Processed’ folder. These two folders, linked together, will become the ‘content information’ of your ‘archival information packet.’ The processed folder will eventually serve as the content information of the ‘dissemination information packet.’
- For each preservation action you undertake on the content information or any part of it (e.g. migration to a new format, renaming, deleting, or reorganizing files), keep track of your actions at an aggreate level in. Describe the actions you undertook in a short narrative statement, and keep this information in text file named “PreservationActions.txt” in the root of the ‘Metadata” folder. This record serves as part of your preservation description information (PDI). (Unfortunately, there are no easy to use tools currently available to track file level peservation actions taken outside of systems such as RODA, etc.) You might consider saving log output, if available, from your conversion applications and putting it in the Metadata folder.
- Run a batch FITS process and an MD5 checksum generator against the processed folder, and put the results into your Metadata folder.
- For each ‘archival information packet,’ use your existing catalog system to track collection-level descriptive metadata. Track AIPS at aggregate level (don’t worry about single files) before doing detailed descriptive records for part of the AIP. Detailed description of the collection can be done using standard archival cataloging tools.
- As a “dark archives,” of the AIP, store all e-records in one folder, containing the subfolders for ‘Metadata’ and ‘Content’, for each set of records you are preserving . The main folder must use a unique identifer that is the same as the unique identifier for the associated descriptive information in your collections management software. This serves as rudimentary packaging information to link the content information, descriptive information, and preservation description information into a complete ‘archival information packet.’
- If posible, provide for redundant file storage and recovery of the AIP) using ZFS (you’ll need tech support) or an online service such as I-Drive). Make sure your descriptive information is similarly backed up.
- Provide access to the ‘processed’ folder, which is a Dissemination Information Packet (DIP). Create a zip file including each ‘processed’ folder and a copy of the collection description/finding aid. Put copies on your website (provide links form your on descriptive information, if applicable) where copyright and other rights considerations allow, and link to this in you cataloging application or finding aid. Provide users an easy method to request and access materials that cannot be placed on line due to copyright or other rights considerations.
It is my opinion that the above system represents core elements of a very rudimentary OAIS system (without sophistacated elements like access controls, etc) , and it can be implemented by most archivists using exisiting tools.