Share via

Azure SQL Managed Instance database not visible in sys.databases but name still appears to exist

João Paixão 0 Reputation points
2026-04-14T16:04:38.1133333+00:00

We ran into an issue on Azure SQL Managed Instance after a database drop operation.

The database no longer appears in sys.databases, so from the normal catalog view it looks like it has already been removed. However:

  • CREATE DATABASE using the same name fails with “database already exists”
    • restoring a database to that same name also fails with the same message
    • renaming another database to that name also fails
    • DROP DATABASE for that name says the database does not exist

This makes it look like the database is both gone and still present at the same time.

Has anyone seen this on Managed Instance? Is this usually caused by an internal drop operation still being in progress or stuck in the backend, leaving the database name reserved even though the database is no longervisible?

The DMV we are looking at is sys.dm_operation_status in master, since it seems to show managed operations such as create, restore, and drop. I’m trying to confirm whether this is the right way to diagnose the issue, and whether there is any customer-side resolution or if Microsoft support is typically required to clear the reserved name.

Azure SQL Database

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. Manoj Kumar Boyini 13,850 Reputation points Microsoft External Staff Moderator
    2026-04-20T14:14:54.2866667+00:00

    Hi João Paixão,

    Our PG team identified that the database drop operation had become stuck in a backend state during the storage file deletion phase. As a result, the database was no longer visible in sys.databases, but its name remained internally reserved, which caused subsequent operations using the same name to fail.

    This condition was due to residual database files that were not successfully removed during the drop operation because of an underlying file path mismatch at the storage layer.

    The backend cleanup has now been completed, including removal of the remaining database files and updating the internal state of the managed instance to progress past the stuck drop operation.

    You should now be able to reuse the database name for create or restore operations as expected.

    Please let us know if you continue to experience any issues.

    1 person found this answer helpful.

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.