HTTP 422 Unprocessable Entity returned by the Quality API during load testing.
We are testing the export limits for quality evaluations using JMeter 5.6.2. The script hits the /api/v2/quality/evaluations/export endpoint with a batch of 500 IDs. It works fine with 50 threads, but fails immediately when we scale to 200 concurrent threads. The response body contains a generic validation error. Is this a rate limit issue or a payload size restriction? Our instance is in us-east-1. We need to know the safe concurrency threshold for this endpoint.
The root of the issue is likely a mismatch between the export payload structure and the backend processing limits for bulk operations. When scaling to 200 concurrent threads, the QAPI endpoint may struggle to parse the evaluation_ids array if the request headers lack proper Content-Type: application/json specification or if the date_range filter is too broad.
Check the export_request object in your JMeter script. Ensure the status field is explicitly set to completed rather than leaving it null. The dashboard often defaults to completed, but the API requires explicit filtering for large batches. Also, verify that the language parameter matches the instance configuration. If the 422 persists, try reducing the batch size to 100 IDs per request. The Performance Dashboard handles this granularity better, and the API may have a hidden limit on simultaneous evaluation exports. This approach aligns with standard enterprise load testing practices for Genesys Cloud.
While the header checks mentioned above are valid, the core issue with a 422 on bulk exports is often the payload structure exceeding the API’s internal validation thresholds for concurrent requests. In my experience managing high-volume schedule publications for our Chicago team, we see similar failures when the request body doesn’t explicitly define the export_format or when the evaluation_ids array exceeds 1,000 entries per call. The Quality API enforces stricter validation under load to prevent backend timeouts.
Try breaking the batch into smaller chunks and ensuring the JSON payload explicitly includes the format field. Here is a robust payload structure that typically bypasses the 422 error:
{
"format": "csv",
"evaluation_ids": [
"eval-id-1",
"eval-id-2",
// ... up to 500 IDs
],
"include_transcripts": false
}
Also, verify that the Content-Type is strictly application/json. If you are hitting this with 200 threads, consider implementing a retry mechanism with exponential backoff in JMeter. The Quality API has a soft limit on concurrent export requests per tenant. If you hit that limit, it returns a 429, but if the payload is malformed due to race conditions in thread generation, it defaults to 422.
We resolved a similar issue in our WFM schedule publishing workflows by serializing the export requests. Instead of firing 200 threads simultaneously, we staggered them using a constant throughput timer in JMeter. This keeps the load steady and prevents the validation engine from rejecting payloads due to temporary state inconsistencies. Check your JMeter thread group settings to ensure you aren’t creating duplicate evaluation IDs in the batch, which also triggers validation failures.