Role explosion happens when every project, team, and exception creates more roles, more rules, and more chaos. In Databricks, what begins as a clean role-based access control (RBAC) plan can swell into thousands of granular permissions. Each new role adds operational weight. Multiply this across large-scale environments, and access control becomes its own bottleneck.
The warning signs are clear. Queries take longer to authorize. New users wait days for the “right access.” Security becomes brittle because permissions drift without notice. The bigger the platform footprint, the more time your best engineers spend on permission mapping instead of delivering value.
The root cause is often a mismatch between how RBAC is designed and how large organizations actually work. Data teams experiment. They create and discard projects quickly. Mappings between users, groups, and roles accumulate but rarely get cleaned up. Overlapping policies create uncertainty, which drives more roles, not fewer.