Need some troubleshooting help with a recurring failure in our bulk export workflows for legal discovery requests. We are utilizing the Genesys Cloud BYOC (Bring Your Own Cloud) architecture with an Edge instance deployed in eu-west-1. The issue manifests when triggering a bulk export job via the /api/v2/recordings/bulk endpoint. While the job status initially transitions to ‘processing’, it fails after approximately 45 minutes with a generic ‘INTERNAL_ERROR’ status in the job details. The specific error code returned in the response payload is ‘BULK_EXPORT_S3_PUSH_FAILURE’, which is not documented in the standard API reference for common export errors. This is critical as we have a strict chain of custody requirement for all digital channel recordings, and these failures are causing significant delays in our legal hold processes.
The environment consists of Genesys Cloud version 2023-10-19, with our Edge instance running version 2023.10.1. The S3 bucket policy allows full write access to the IAM role associated with the Edge, and manual uploads via AWS CLI work without issue. We have verified that the KMS key encryption settings match between the Edge configuration and the S3 bucket defaults. The recordings in question are primarily from digital channels, specifically WebRTC softphone sessions and SMS transcripts. Interestingly, PSTN call recordings exported via the same job configuration succeed without error. This suggests the issue may be related to how the Edge handles metadata serialization or payload size for digital channel artifacts before pushing to the external storage provider.
We have attempted to increase the timeout settings in the Edge configuration and reduced the batch size of recordings per job, but the failure persists. The audit trail shows the Edge successfully retrieves the recording data from the Genesys Cloud core but fails during the transmission handshake with S3. We are looking for insights into whether this is a known limitation with BYOC Edge handling large JSON manifests for digital channels or if there is a specific header requirement we are missing in our S3 integration setup. Any guidance on debugging the internal Edge logs for S3 push failures would be appreciated.