Every reorg creates orphans. Teams get dissolved, merged, or moved, and the databases they ran do not get formally handed to anyone. The quiet work the old team did just stops. Nobody decides to stop it. It stops because the person who used to patch, check the backups, and watch the monitoring now sits in a different cost centre with a different remit.
I have run the team that absorbed someone else’s systems. You inherit a folder of servers and a vague instruction to keep the lights on. Two of them you understand. The third is a name on a list that nobody in the new structure can explain. It still runs. It still serves an application somewhere. And it has been months since anyone confirmed its backups restore, because the person who used to confirm that left in the shuffle.
The danger is that an orphaned database does not look broken. It looks fine, right up to the moment it is not. Patching lapses because no named person owns the maintenance window. Monitoring alerts route to a distribution list that was deleted in the move. A disk creeps toward full and the warning lands in a mailbox nobody reads. The server has been telling someone the whole time. The shuffle just removed the someone.
Finding the orphans is a paper exercise before it is a technical one. Take the full list of production databases. Next to each one, write a single human name. Not a team. Not a queue. A person who would take the call at the wrong hour. The databases where you cannot fill that box are your orphans. Then check the basics on each. Is it patched. Are backups running and tested. Does monitoring point at an inbox a real human reads. That gap between the list and the named people is the whole risk, sitting in plain sight.
The fix is not technical either. It is a decision. Each orphaned database gets one named owner with the time and access to actually run it, written down where the next reorg cannot quietly erase it. If a system genuinely has no business owner left, that is worth knowing too. It might mean the system should be retired. An unowned production database is a far better thing to decommission on purpose than to discover during an outage.
If your org has reshuffled in the last year or two and you are not certain every production database has a name beside it, that is the moment to look. The free 15-minute health check is read-only. It runs against your estate, touches nothing, and produces a graded plain-English report you can put next to that ownership list. It will not tell you who should own each database. It will tell you, in plain terms, which ones nobody has been looking after.
Want to know if this is sitting in your estate? We run a read-only check and hand you a graded report in plain English.
Get your free health check