Thanks for sharing what resolved the issue. In case that’s unclear to anyone, see the resolution:
One doesn’t have to recreate the database when migrating to a service. One can move it instead.
One can also recreate the database without migrating, and it might be a simpler experiment to try.
But there are a number of experiments suggested, and it would be good to gain an understanding.
One possible way to get both might be to time the Verify files
, move database aside, Repair
,
then time the Verify files
again. Checking Profiling log for SQL execution times may also help.
A recreated database will likely be smaller than even the original after vacuum, because it will lose
some history such as the job logs and remote operations done. log-retention keeps them trimmed,
and I can’t think of any data-oriented tables that grow to the sky. Still, it might be fast after recreate.