Azure Document Intelligence completely hanging - processes never complete, requiring manual termination after minutes

Question

Issue Summary

Our production Azure Document Intelligence service is completely hanging during PDF processing. Documents that previously processed in <1 second now hang indefinitely for minutes, never completing, forcing us to manually terminate requests.

Environment

• Service: Azure Document Intelligence (prebuilt-read model)
• Endpoint: West Europe region
• Document Type: Medical journal PDFs (1-5 pages, containing scanned text images)
• Output Format: markdown
• API Response: 202 Accepted, but polling never completes

Timeline

• Previous performance: <1 second processing time
• Current behavior: Hangs indefinitely - no completion after 2+ minutes
• Started: September 23, 2025
• Frequency: Every document, 100% failure rate

Technical Details

• No rate limiting headers in response (confirmed not quota issue)
• API calls return 202 status successfully
• Polling phase hangs completely - poller.pollUntilDone() never returns
• We've waited 2+ minutes with no completion
• Must manually terminate/timeout requests
• No error messages - just infinite polling
• Same PDF documents that processed successfully before

Business Impact

• Complete production system failure
• Zero successful document processing
• Healthcare application completely unusable
• Customer operations at standstill

Code Behavior

// This call hangs indefinitely - never resolves
const result = await poller.pollUntilDone();
// After 2+ minutes, we terminate manually

Questions

1. Is there a complete service outage in West Europe?
2. Are other customers experiencing total Document Intelligence failures?
3. This is a production emergency - what immediate steps can we take?
4. Should we immediately failover to a different region?

Answer 1

Hi Anders Myrmel,

We can confirm that there was a service impact affecting Azure Document Intelligence in several regions. The engineering team identified the cause and has now fully mitigated the issue.

At this time, service functionality should be restored. If you are still experiencing errors, please try your request again and monitor the Azure Status Page for any real-time updates.

Check the Azure Service Health dashboard for the latest updates on your region.

We sincerely appreciate your patience and understanding while this was being resolved.