Skip to main content

Integration Alert: Troubleshooting Your Data Sync or Export Disruption

Recurrency will provide alerts when we detect issues with your integration. This document will outline common errors and how to resolve.

Updated over a week ago

When Recurrency has detected a disruption in your integration, you're admins will receive an alert to their email. This alert means that we were unable to successfully complete a scheduled data sync or export job from your ERP or database system.

While we monitor for failures automatically, most disruptions occur due to configuration or access issues on your system — such as expired credentials, unreachable endpoints, or missing data.


Common Root Causes

Below are common scenarios depending on your integration type:

Credential or Token Expiry

  • API tokens may expire

  • Database user passwords may change or rotate

  • IP restrictions, firewall changes, or issues with the Recurrency integration agent may block access

ERP System Unavailability

  • System downtime or maintenance during the sync window

  • API endpoints not responding or returning unexpected errors

  • ERP service layer restarted or inaccessible

Configuration Issues

  • Schema or field changes not reflected in Recurrency’s integration

  • Permissions removed for Recurrency’s API or DB user

  • Feature-specific prerequisites (e.g., Min/Max exports) misconfigured

Alert Types

Data Sync

If you receive this alert, then Recurrency has been unable to connect to your ERP to read new data, or export data back, to the ERP.

  • Data will be stale since the last time Recurrency could read data from your ERP

  • Exports back to the ERP will likely fail and need to be retried

If you are using an on-prem ERP instance, you likely need to reinstall the Recurrency Integration Agent to re-establish connection.

SISM Export Issues

If you are using Epicor Prophet21, we default to using SISM to export Purchase Orders from Recurrency to your ERP. SISM runs as a subprocess in your ERP, so it can often get stuck or otherwise be too slow to respond to requests.

If you receive this alert, you need to restart SISM.

If you are consistently receiving this error, you can reconfigure SISM to see if that resolves the issues going forward.

Did this answer your question?