Pushed the custom edge routing module to client staging org using genesys-cloud-sdk-js@3.97.4. The /api/v2/edge/byoc/networks/{networkId}/validate endpoint doing jack all. It’s rejecting the POST with a 400 Bad Request right after the handshake completes. The payload returns a clear error about VPC endpoint service mismatch, but the AWS console shows the endpoint sitting perfectly in us-east-1. IAM role has the ssm:GetParameter and ec2:DescribeVpcEndpoints policies attached. Checked the trust policy three times. Everything looks correct on that side.
Console logs show the SDK sending the request with the correct bearer token. The response headers point to the edge gateway routing layer dropping the call. Switched the target region to us-east-2 to check if the validator was holding onto old cache data. Same 400 response. The network CIDR block matches what the client’s firewall team approved last week.
Architect flow on the GC side is completely idle during this step. The integration just hangs at the deployment stage. Client’s AWS ticket says the VPC peering is active and healthy. Maybe the BYOC validator is checking a deprecated IAM path or something. The docs mention a region alias mapping but that feature got deprecated in the Q3 release notes.
Here’s the raw response body coming back:
{
"errors": [
{
"code": "invalid_request",
"message": "VPC endpoint service does not match the expected AWS region or IAM role lacks ssm:GetParameter permission"
}
],
"requestId": "req_8f9a2b1c-4d5e-6f7g-8h9i-0j1k2l3m4n5o"
}
Network ID is net_92837465. SDK logs show the payload matches the schema exactly. Retried the call with a fresh OAuth token generated five minutes prior. Still hitting the same wall. The validation step keeps timing out at exactly 12 seconds.