You could spool objects that were invalid before the migration. Then go about attempting to make them invalid after the migration. But again, many of these objects will automatically recompile when accessed. With triggers, you can affect the outcome with a "alter trigger blah disable;".
Originally posted by Axr2 Yeah, I'm guessing that he was concerned about invalid triggers essentially. The utlrp would then forcibly make it valid afterwards, even if someone had explicitly had them down as "invalid" before.
If they can be valid. There might be a case where the upgrade is causing packages to not be valid.