Skip to content
DevOps AI ToolKit
Newsletter
All prompts
AI for MySQL Difficulty: Intermediate ClaudeChatGPT

MySQL Foreign Key & Data Integrity Audit Prompt

Audit a MySQL/MariaDB schema for missing or mismatched foreign keys, orphaned rows, ON DELETE/UPDATE behavior risks, and constraint gaps, then propose a safe plan to add integrity without breaking writes or replication.

Target user
DBAs and backend engineers
Difficulty
Intermediate
Tools
Claude, ChatGPT

The prompt

You are a senior MySQL DBA auditing referential integrity on a schema that has grown organically, where some relationships are enforced and others are only implied by the application.

I will provide:
- `SHOW CREATE TABLE` for the related tables (parent + children)
- Existing FK definitions from `SELECT * FROM information_schema.referential_constraints` / `key_column_usage`
- Engine (must be InnoDB for FKs), and whether tables are partitioned (FKs unsupported on partitioned tables)
- Approximate row counts and whether orphan rows are suspected
- Write volume and whether the app currently relies on application-level integrity

Your job:

1. **Map implied relationships** — identify columns that look like foreign keys (xxx_id) but have no constraint, and check column type/charset/collation match between parent PK and child column (a common reason ADD FOREIGN KEY fails).
2. **Detect orphans** — give the exact LEFT JOIN ... WHERE parent IS NULL queries to count orphaned child rows that would block adding a constraint.
3. **Recommend ON DELETE/UPDATE actions** — weigh CASCADE vs RESTRICT vs SET NULL per relationship, warning where CASCADE could trigger large unintended deletes or replication surprises.
4. **Plan safe rollout** — clean orphans first (in batches), then add constraints, ideally via ALGORITHM=INPLACE or pt-online-schema-change to avoid long locks on big tables; note that adding an FK still validates existing data.
5. **Decide enforcement boundary** — where FKs are impractical (partitioning, sharding, very hot tables), recommend application-level checks plus periodic reconciliation queries instead.

Output as: (a) relationship map with type/collation mismatches, (b) orphan-detection queries, (c) recommended actions per FK, (d) ordered rollout plan, (e) where to keep app-level integrity.

Advisory only: present queries and DDL for review; never auto-run CASCADE-bearing constraints or orphan deletions, and require a backup before integrity changes.
Newsletter

Free: the DevOps AI Incident-Triage Cheat Sheet

Subscribe and we’ll send you the one-page cheat sheet — plus weekly AI prompts, automation ideas, and tool reviews for infrastructure engineers. One email a week. No spam, unsubscribe anytime.

  • AI Incident-Triage Cheat Sheet (PDF)
  • Access to 2,104 DevOps AI prompts
  • One practical workflow email per week