---
title: Validation & Testing
slug: validation-and-testing
createdAt: 2025-10-29T21:30:32.862Z
---

Thorough validation ensures each connector behaves predictably and avoids unintended bulk changes.

# Pre-deployment Checks

- Confirm the **schema** matches the source data — all required columns exist and are typed correctly. 
- Validate **key fields** for uniqueness in the source. 
- Review **delete strategy** (flagged vs. result-set) to confirm it aligns with data behaviour. 
- Ensure **history logging** is enabled for traceability. 

# Dry Runs

1. Set processImport/maxChanges and processImport/maxDelete to low values (e.g., 50). 
2. Import a small data subset. 
3. Verify that detected changes appear in the **connector data table** with expected Action states. 
4. Review **pending changes** in the web interface before running the process step. 

# Functional Testing

- Validate that **creates, updates, and deletes** trigger the correct event handlers. 
- Check that scripts, job submissions, and notifications execute as expected. 
- Confirm that multi-value and foreign key columns are parsed correctly. 
- Review that role or user links resolve properly via the relevant link handlers. 

# Error Testing

- Simulate missing or malformed data to confirm retries and error handling work. 
- Verify that email notifications fire on error conditions. 
- Confirm that connectors stop gracefully when exceeding maxDelete or maxChanges thresholds. 

# Post-deployment Monitoring

- Review **history tables** after each run to confirm change tracking. 
- Monitor **connector logs** for warnings or retries. 
- Spot-check **sample rows** across the import, data, and history tables. 
- Schedule automated notifications on failure conditions.