Architect BYOC Trunk Selection Logic Failing

Architect & Flows
1

Struggling to understand why the Architect flow fails to route calls to the correct BYOC trunk when using the ‘Select Trunk’ action. We have 15 trunks across APAC regions. The flow uses a decision step based on the dialed number prefix, but calls are randomly distributed instead of following the defined failover order. The Genesys Cloud version is 2024.1. Is there a known issue with the trunk selection logic when multiple trunks share the same outbound route pattern?

2

This is caused by…

Cause:
The ‘Select Trunk’ action in Architect does not inherently enforce a strict priority order when multiple trunks match the same outbound route pattern unless explicitly configured within the action’s properties. By default, the system may distribute calls based on available capacity or round-robin logic across matching trunks, ignoring the intended failover hierarchy. This behavior is documented in the Genesys Cloud release notes for version 2024.1, where enhancements to trunk selection logic were introduced but require explicit configuration to override default distribution methods.

Solution:
To ensure calls follow the defined failover order, you must configure the ‘Select Trunk’ action to use a specific trunk list with priorities. Here is how to adjust your flow:

  1. Identify Matching Trunks: Ensure all 15 trunks are correctly mapped to the outbound route pattern in question.
  2. Configure ‘Select Trunk’ Action:
  • In the Architect flow, add a ‘Select Trunk’ action.
  • Under the ‘Trunk Selection’ settings, choose ‘Specify Trunks’.
  • Add each trunk to the list in the desired failover order.
  • Set the ‘Failover’ option to ‘Try next trunk if unavailable’.
  1. Validate Route Patterns: Confirm that each trunk has a unique or correctly prioritized route pattern to avoid ambiguity.
  2. Test the Flow: Use the flow testing tool to simulate calls with different dialed number prefixes and verify that the correct trunk is selected based on the defined order.

Additionally, review the Performance Dashboard metrics for each trunk to monitor call distribution and ensure the failover logic is functioning as expected. Adjust the trunk priorities if necessary based on real-time performance data.