DataGrip and Database Server: Read-Only Mode Enforcement Boundaries
0 reputation · 27 Aug 2020, 04:07 UTC
Client-Side Restrictions vs Server-Side Permissions
DataGrip includes a 'Read-only' toggle within the Data Source properties designed to prevent the execution of DML and DDL statements. This feature acts as a safety layer to avoid accidental data modification in sensitive environments.
While this setting restricts the IDE from sending destructive commands, it operates independently of the database server's Role-Based Access Control (RBAC). There is an unresolved distinction in how the IDE handles the intersection of these two layers, specifically when a user has write permissions at the server level but has enabled read-only mode within the client.
If a connection is marked as read-only in DataGrip, but the underlying database user possesses administrative privileges, the restriction remains a client-side preference rather than a hard security boundary.
- Does DataGrip provide a mechanism to synchronize the IDE's read-only state with the actual session permissions of the database server?
- How does the IDE handle scenarios where server-side permissions are more restrictive than the client-side configuration?