Stuck on a bulk export job that refuses to transition from pending to processing when filtering for specific digital channel interactions.
We are attempting to pull a complete chain of custody dataset for a legal discovery request involving WhatsApp and Facebook Messenger interactions from the last fiscal quarter. The environment is Genesys Cloud PureCloud, and we are using the standard Bulk Data API (v2) to trigger the export. The job is configured to push directly to our designated S3 bucket, which has been verified with correct IAM permissions and CORS settings. The filter query is constructed as interaction.type=webchat OR interaction.type=whatsapp with a date range of 2023-01-01T00:00:00Z to 2023-12-31T23:59:59Z.
The issue manifests immediately after submission. The job ID is returned successfully with a status of PENDING. However, after 45 minutes, the status remains PENDING with no progress indicated in the percentComplete field. No error message is attached to the job object itself. Checking the S3 bucket, no prefix folder has been created, suggesting the export engine has not even attempted to write the initial manifest file. This is critical for our audit trail compliance, as we need the raw JSON files with full metadata including user.id, interaction.id, and timestamp to maintain the integrity of the discovery request.
We have confirmed that smaller date ranges (e.g., single week) complete successfully, so the S3 connection is valid. The problem appears specific to the volume of digital channel data or perhaps a timeout in the query parsing stage for these specific interaction types. The API logs show no 429 or 500 errors during the initial POST request. Is there a known limitation on the number of interaction types combined in a single filter for bulk exports, or should we be splitting these into separate jobs per channel type to avoid this hanging state?