Issue
When using Segment Reverse ETL to sync records to Adobe Campaign Standard (ACS), you may experience intermittent 500 Internal Server Errors on the ACS side. These errors are not visible in your Segment logs but are observed in ACS, often during scheduled sync windows.
Common error messages include duplicate key constraint violations and connection issues such as "no route to host". You may initially suspect API call rate limits or throughput constraints as the cause.
Product
Twilio Segment
Environment
Segment Console
Cause
The primary cause of the 500 Internal Server Error is a duplicate key constraint violation in the ACS database. This occurs when records with the same key are sent multiple times, resulting in a conflict.
Connection errors may also be related to your ACS or Azure PostgreSQL configurations, such as connection pool limits or network setups. Segment logs do not show evidence of API rate limiting or throughput spikes during the affected sync windows.
Resolution
- Review the ACS Database: Ensure records are not being sent with duplicate keys, which causes constraint violations.
- Check ACS/Azure PostgreSQL Configuration: Review your connection pool limits and network settings to address any "no route to host" connection errors.
- Monitor the Sync Overview: Use the Segment Console’s Sync Overview tab to monitor retry events and API call volume for each specific sync.
- Contact ACS Support: If errors persist, coordinate with ACS support to further investigate your database constraints and network configuration.
Additional Information
- Segment implements retry logic for destination API calls, which is visible in the Sync Overview tab.
- No Segment-side throttling or unusual API spikes were observed in the referenced sync logs for this issue.
- Scheduled syncs typically run every 6 hours; monitoring API traffic specifically during these windows can help you identify error patterns.