Debugging And Error Recovery

Debugging and Error Recovery (DBA)

Overview

Use a structured workflow to reproduce, isolate, and fix migration/schema failures with minimal blast radius.

When to Use

• Migration failures
• Constraint/index behavior regressions
• Query-plan regressions tied to schema changes

When NOT to Use

• New feature implementation with no active failure to diagnose

Inputs

• Error output and failing migration/query context
• Relevant schema objects and migration history

Process

1. Reproduce issue in controlled steps.
2. Isolate failing object/boundary (table, index, constraint, order).
3. Implement minimal corrective change.
4. Re-verify migration path and integrity behavior.
5. Document root cause and mitigation.

Verification Checklist

• Failure is reproducible before fix
• Root cause is identified, not guessed
• Fix is minimal and targeted
• Validation confirms issue resolved without regressions

Failure Modes to Avoid

• Broad speculative schema changes
• Skipping root-cause analysis
• Fixing symptoms while leaving migration path fragile

Output Requirements

• Use summarize for final output.
• Include root cause, corrective action, and validation evidence.