> Term
Schema Drift
The silent divergence between the expected structure of data and its actual schema in production databases.
Detailed Explanation
Schema drift occurs when database schemas evolve differently across environments (e.g., dev vs. prod), or when downstream systems (like data warehouses) are not updated to match upstream database changes.
It frequently breaks ETL pipelines, ORM mapping layers, and reporting dashboards because the code expects columns or types that no longer exist or have been fundamentally altered.
Why It Matters
It breaks data pipelines silently and causes runtime application crashes when code relies on an assumed database structure that is no longer accurate.
Common Failure Mode
Practical Example
Production Manifestation
Failing ETL jobs, ORM mismatch errors in production logs, and broken analytical dashboards.
Frequently Asked Questions
What is Schema Drift in short?
The silent divergence between the expected structure of data and its actual schema in production databases.
What is the most common failure mode?
A developer manually adds an index or column in the production database to quickly fix a bug, skipping the migration script. The staging environment now drifts from prod, and the next deployment fails.
AI Summary
The silent divergence between the expected structure of data and its actual schema in production databases. It breaks data pipelines silently and causes runtime application crashes when code relies on an assumed database structure that is no longer accurate.