I have a couple of folders with several thousand product items in each. All the products are using the same data type.
Recently I had to add a new field to the type and attempted to deploy this to my DEV cloud environment. This failed multiple times which appeared to be a problem of db timeouts.
I ended up deleting all the products and clearing the Trash, and then the update succeeded. (Then I had to repopulate all the products. This was ok since there is a batch job that does it.)
When I deployed the DEV up to PROD I made sure to delete all products prior to beginning the deployment. That deletion process took over an hour all on its own. Clearing the Trash took another hour. A deployment process that normally takes about 20mins ended up taking nearly 4 hours.
So is there a better way to do this - updating the schema of a data type which is used by thousands of items?
The problem is mostly that it just takes a really really long time with no indication of progress up front. Deleting a folder of items when there are over 1000 entries also takes a really looooong time. And then I had to also delete those same items a second time from the Trash, else the schema change still took a loooong time.
And this was just an addition of a new field, not even a modification. (In SQL adding a field is usually quick, though modifying an existing one requires creating temp tables and migrating data, so I understand that that takes longer.) There are not 10s of thousands of records, just a few.
Does anyone have advice on a better strategy for mods to a schema when there are lots of records?
Cheers,
AJ