Hi,
I am investigating upgrading my University’s DSpace instance from 5.6 to 7.6.1, moving to a new server in the process, following https://wiki.lyrasis.org/display/DSDOC7x/Migrating+DSpace+to+a+new+server. Starting with little knowledge of DSpace, both server- and user-side, I have been able to get 7.6.1 up and running as a fresh install, and have moved on to dumping/restoring the database etc.
After successfully restoring the database from a dump and running “dspace migrate ignored”, both “dspace index-discovery -b” and “dspace oai import” fail with “java.lang.IllegalArgumentException: More than one result found” and a stack trace resembling what is reported (“Another question”) in https://groups.google.com/g/dspace-tech/c/DIIDMcoWSzk/m/JPvN1E_qBgAJ.
In my case, “select item_id, COUNT(item_id) from workspaceitem group by item_id having count(item_id)>1;” yields 71 duplicated IDs, mostly with a count of 2 and a handful with 3. In most cases, the workflow_item_id is the only difference between the duplicated rows but there are some with various other differences as well. I assume that ultimately the only solution is to delete the duplicates, as mentioned in the above link, but how should I decide which one to keep (and what other implications are there)?
Thanks,
Alex