Hello -
In our current (ArcMap/GN) implementation we have a process that runs nightly uses version differences to detect any features that have been deleted from SDE.Default in the previous day. Any features found are written to a "DeleteHistory" file geodatabase. Features are never removed from the DeleteHistory database.
I won't enumerate, but there are all kinds of cases where a feature/features can get unintentionally deleted. The process has certainly saved my bacon.
It seems like branch versioning will retain history of what has been deleted... up until the point where we need to perform a "prune". At least it seems so. My understanding is that "prune" will remove any archived features no longer associated with a version -- meaning it will remove rows that would otherwise describe what was deleted.
If this is true, has anyone developed processes to retain delete history beyond the "prune" operation? We're thinking that this *might* be accomplished simply by creating a file geodatabase snapshot before the prune occurs.. though, if we have to search back over multiple "prune" operations we would have to search over multiple FGDB snapshots.
Any thoughts or suggestions would be much appreciated.
Thanks,
Ed
FWIW - We've decided to continue our current Delete History detection process going forward with the UN. Even though we could probably find these in the branch versioned tables, the combination of having the information in one known place, plus not needing to worry about the effects of pruning make us feel best about keeping the same process.
Ed