2023-01-09 - Data Migration Subgroup Agenda and Notes

Darsi/Stanford: Ready to build Nolana and do a test load.  Working on extracting reference data so can reload it.  Ready to try loading boundwiths and merge marc holdings into FOLIO holdings.

Circ data: old loans for lost items/really old fines.  Consensus in group is not to load these.

Deleting inventory and reloading? What about loans on those items?  Ian: if UUID for item is deterministic it should link back up when you reload.

Ingolf: needs to load some marc data.  what endpoint should he use to load? Ian: want to use endpoint that will load SRS (the marc) and then use default mapping to create instance.  But note that this will not create deterministic instance UUID.  mod-inventory-storage.  EBSCO FSE FOLIO migration tools to load inventory are good (#FSE_folio_migration_tools slack channel) https://github.com/FOLIO-FSE/folio_migration_tools

Texas State will migrate to FOLIO (ERM and Acq tentatively August 2023, the rest in 2024). EBSCO will host and support implementation/migration.  Thoughts so far on data migration: not migrating really old/lost loans, old fines.  Users: University uses Banner, auto downloads

Advice:

• (Darsi) : load default reference data. Keep UUIDs persistent on each FOLIO instance.
• (Ian) : user input, how to maintain source of truth from campus systems; what will be ID that is “link” for updates.  Plan reporting solution early on.  LDP/Panorama/LD Light(?).  Have a good representative inventory/bibliographic testing chunk (something in 10s of 1000s) that you can tweak, delete, reload multiple times. Deterministic UUIDs for items helps a lot for those with transactions.
• (Ingolf): plan to be able to erase data and reload. A way to resume from point of error

• documentation, then maybe training – if this makes sense
  • where would it go?  installation info is already in docs.folio.org, so an overview page for data migration might be ok there.  wiki for sure would make sense