I have an OOB system that has hundreds of tables. Each table has a ROWSTAMP column (integer) that gets populated by a trigger — each time a record is created or updated. All the tables/triggers use the same db sequence. Each time a record is updated anywhere in the db, that one sequence is used to get the next available integer. (I suppose I could consider the sequence to be a sort of global sequence).
IBM: Usage of ROWSTAMP column in Maximo tables
It’s unfortunate that those ROWSTAMP columns are integers, and not dates. What I really need in the tables are CHANGEDATE columns. Dates would be useful for data entry troubleshooting, reporting, and analysis in general. But I can’t really justify customizing hundreds of tables and adding hundreds of triggers in the OOB system. The customized tables wouldn’t be supported by IBM and it wouldn’t be worth the effort/added complexity.
I feel like those ROWSTAMP columns are so close to being useful to me. I can’t help but wonder:
Is there a way to retrieve the date that those ROWSTAMP sequence values were generated?
For example, and I doubt this is possible, could I add a DATE column to the sequence, and then join from my table to the sequence to get the date? Or could I use some sort of logging mechanism on the sequence to put rows (with dates) in a table whenever the next sequence value is used?