Comment 2 for bug 1496747

I think we should probably stick with the use of oslo.db's soft delete and update the schema to allow this. Interesting that they set that to the id though. Not sure the reason for that.

It might be good to get other's opinion on this as well.

And it might be good to review all our table schema's and see if there are other instances where this could be an issue. Luckily I doubt there's anyone who currently uses anywhere near 128 volume types.