EPrints Technical Mailing List Archive

See the EPrints wiki for instructions on how to join this mailing list and related information.

Message: #08365

< Previous (by date) | Next (by date) > | < Previous (in thread) | Next (in thread) > | Messages - Most Recent First | Threads - Most Recent First

Re: [EP-tech] Trying to understand a disaster after a database update

Hi Laurent,

So to confirm: Are there are no records in the eprint database table or you just don't see any records when you look at the Manage Deposits page in the repository?  When you ran recommit at the end of your set of instructions.  How long did this take to run and how many record did you have before you started this whole process?  If it took quite a long time then it suggests some how the recommit process caused you problem.  However, if it ran quickly this suggests you eprint records were already deleted or broken so could not be displayed in the Manage Deposits page.

As a general comment on your solution.  I would have not dumped out and reimported database tables.  I would have created a rule (block of code) in your archive's cfg/cfg.d/eprint_fields_automatic.pl to map (probably just copy) the old field to new field.  You would have still had to have the old field defined in eprint_fields.pl but as long as you had removed the old field from the eprint workflow and any other config files (except eprint_fields_automatic.pl and eprint_fields.pl) it should not be editable, or visible to non-logged in users.  You would then need to run epadmin recommit like you did in your original instructions and once this completed you could remove (probably best to comment out) references to the old field and the rule to map the old field to the new field.

Regards

David Newman

On 16/11/2020 10:49, Laurent Cloarec via Eprints-tech wrote:
CAUTION: This e-mail originated outside the University of Southampton.
Hello everyone

The initial problem : many (about 1,000) metadata values entered by archive editors in a non appropriate field (used in parallel for another purpose by another import mechanism), let's say "infoX". Into the database, this information was stored in a table looking like "eprint_creators_infoX(eprintid,pos,creators_infoX)"

The solution employed :
  1. create a new metadata/field (let's say "infoY") among the creators metadata (with "eprints_fields.pl") and update the database structure in consequence => creation of a new table "eprint_creators_infoY(eprintid,pos,creators_infoY)"
  2. backup (SQL values dump) the important values previously entered and stored into the "eprint_creators_infoX" table;
  3. from this backup, create a new SQL file where the "infoX" column name is replaced by "infoY";
  4. import these values into the new table ""eprint_creators_infoY";
  5. delete the previously entered value from  "eprint_creators_infoX" table (there was an objective criteria to distinguish the entered from the imported values);
  6. run "bin/epdamin recommit <archivename> eprint"
The disaster/issue that happened : all the records from "eprint" main table have been deleted, and the archive/repository consequently appears as empty!!!

Could someone explain this???

Best regards
--
Laurent Cloarec
Service Commun de la Documentation - Service du Numérique Documentaire
Université Toulouse 1 Capitole

*** Options: http://mailman.ecs.soton.ac.uk/mailman/listinfo/eprints-tech
*** Archive: http://www.eprints.org/tech.php/
*** EPrints community wiki: http://wiki.eprints.org/

Virus-free. www.avg.com