2019-06-17 Lessons Learned from Recent Migrations Meeting notes

Owned by Karen Newbery
Last updated: Jun 17, 2019
3 min read

Date

Attendees

Discussion items

Chicago and Lehigh

  •  Best things
    •  Reduced loan policies
    • Did not migrate any acq data - kept it in final state along with final copy of Horizon data
      • frequently accessed for reporting
      • batch process created for open orders to be created as new orders in OLE
    • Michelle Suranofsky timed the front-end the same when the back-end changed so users didn't know the difference
    • Data Clean-up projects
      • Circulation
        • Long overdues
        • bills
        • loans
        • things that had workarounds - charging to library departments
      • Cataloging
        • Item Types
        • Locations
        • Shadow records - were they shadowed from migration to Sirsi?
        • Lost/Missing
      • Acq
        • old open orders
        • reviewed funds
      • Current clean-up projects ongoing because data is available differently in OLE
        • logs for capturing errors/data failing to populate en masse - working to normalize errors
        • lack of 33x data is something they want to make use of - enrichment project as they find gaps
        • have distributed technical services with a shared catalog - some errors are from different libraries having different policies. Reducing niche data.
  • Things to watch out for
    • did not find many data migration errors until well into use of OLE
    • loading data over and over and over and running reports against it to find errors
    • Easier to spot bad data when there's fewer possibilities that data could be. ie: reduce number of locations
    • Building in reference table validation into integration tool - has this been shared? Dale Arntson would be who to talk to.
    • Did not retain identifiers for HOL/Items. Did retain Bib identifiers (Chicago). Can run reports against the bib #.
    • Lehigh - migrated serial orders and all history on that order. When adding an order it was a nightmare.
    • Chicago retained some order data in 9xx fields to provide info in case linking on bib id didn't work.

How did receiving manager know what was ordered, when it was due, how much it should cost, all based on history? Map history to a string and attach it as a note?

Find out what your options are for getting data out of your old system.

Chalmers - some things are easier to do manually, so they're going that route instead of scripting. Aimed for getting data out of Intota (ERM) - will be manually migrated due to a clean-up process that's ongoing. Biggest issue for Chalmers is getting their bib data in line with the union catalog (LIBRIS) - lots of clean up. Don't have holdings records in CRM, so the data looked very different. In some cases, they're going to have bad data in FOLIO. Doing some data enhancement in Sierra.

Any problems with bound-widths? Chalmers  - yes. There may be multiple solutions in FOLIO. What she ended up doing was re-cataloging the bound-withs in Sierra. Only had about 30 bound-withs. Issue fixed before migrating data to FOLIO.

Logging during test migrations has been very helpful.

Chicago doing an enrichment project for pre-AACR2 - pulling in an updated record from OCLC.

Thinking about data that could be left behind - Chalmers - note fields have been cleaned up. Fields that don't contain unique data left over from previous migrations. User data (GDPR). Item data that's in the bib record from their union catalog.

Don't sell your staff short - transition may be difficult for a while, but they may be more flexible than we think. We need to be positive that the changes will be better. Be honest with staff that there will be bumps, that it's a paradigm shift, and that we'll all work together to fix those problems.