The recommended approach is still to sequentially update following the steps for each release. That said, there are now some community-developed scripts that may help automate that at https://github.com/gdcc/dataverse-recipes/tree/main/shell/upgrades (thanks to Don Richards, JHU). There is also work being done by Oliver Bertuch on updating through multiple versions at once (I see he has a talk at the Dataverse Community meeting in June - https://web.cvent.com/event/497af056-3d1e-4268-a7b7-02d3e2ba04a2/websitePage:f33ed6ae-4646-48ff-9cf9-f920f0c8c7d5?session=bb57e47b-be9f-48c4-8f81-01e9fd30963e ).
In general, the recommendation to do sequential updates is for two reasons
- simplifying debugging if you miss a step in any given update, and
- because there are some updates where skipping the deployment of the war file for that version will result in a failure (basically when one version adds a database table and a subsequent version modifies it – if the first version is never deployed, the modification fails because the table does not yet exist). If I understand correctly, Oliver’s work is able to handle this case.
-- Jim
--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
dataverse-commu...@googlegroups.com.
To view this discussion visit
https://groups.google.com/d/msgid/dataverse-community/69ae6c88-df20-41b1-b55d-a0f41c398a6cn%40googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/dataverse-community/0fef5561-dbdc-4b55-8b36-cd456100ddabn%40googlegroups.com.