Getting Started with LOCKSS, Midge Coates, 2013-06/Step 1: Decide what you want to preserve.

You’ll need to think about what you want to preserve—the original document files, the “added value” files (transcriptions, etc.), the entire collection in its content‐management software (e.g. a CONTENTdm collection), or something else.

Each Archival Unit (AU) must be Web-accessible so the LOCKSS daemon can get at it. That means that it needs to be put on a Web-accessible server computer. If you have a server, but it isn't Web-accessible or access to it is blocked (at the firewall, for example), the LOCKSS crawl won't work. You can use a firewall to protect your collection, but make sure the LOCKSS daemon has access. Check with your IT support person or system administrator to confirm that this is the case.

In the case of digital collections that use proprietary content-management systems like CONTENTdm, it is advisable to preserve the original image collection (that is, the digital masters or archival TIFs) with the CONTENTdm metadata. You can do this by exporting the metadata from the CONTENTdm collection (or other content management system) as a tab-delimited text file and storing this file in the same folder with your digital master files on the Web server. For this to be most effective, your collection metadata should include identifiers that match up to your original image files.

If you've made transcription files for a collection (audio, handwritten documents, etc.) using MS Word, you should probably re-save these as plain text files. Non-proprietary files are smaller and won't become obsolete.