Flowable 6.6 Upgrade Recovery

My production environment runs on Flowable 6.4, but for development we are running 6.6. Every so often we will pull down our production data and install it into the lower environments, so we are running on the same data that our users work with. This process typically works fine, except that this time we made a mistake. One of the 6.6 sequences got left in and caused the upgrade process to tank (log snippet below has had server info sanitized).

It was an error in our data refresh process that left the sequence in. The question is therefore:

Is there a way to recover/repair/rollback the upgrade process so that when an issue like this does arise, we can simply fix the problem and retry the upgrade?

If not, is there a way to trick Flowable into re-running the upgrade without blowing up all the object creation steps? Or would that result in pain and suffering, meaning it’s easiest if we simply fix our data refresh process (which entails using the Oracle data pump utility, so we are dropping the objects in the schema and then importing the 6.4 objects and data rows).

Sadly, DDL is not transactional (like in postgres, where this would have rolled back and nothing would have happened).

The only thing I can think of is doing the reverse operations as in those sql files that were applied … or upgrading the production to 6.6 ofc.

Good to know for future reference.

I’ll make note that if we do run into an upgrade error like that, all we can do is fix the problem in our refresh process (means we missed an object that 6.6 added that we need to remove) and re-run the process.