I just checked in Sweden central for replicating this issue.
As part of my repro, I uploaded 20 small files (under 100 Kb), It was working.
Suggested steps
- Agree to Azar on using smaller batches and smaller file size.
- Also, would suggest increasing TPM from model deployment for faster indexing.
- Switch to another available region.
- Change to latest GA API version for existing deployments.
Below are Constraints on file size and type.
- The maximum file size is 200 MB. Each file should contain no more than 5,000,000 tokens per file (computed automatically when you attach a file).
- Supported extensions- c, .cs, .cpp, .doc, .docx, .html, .java, .json, .md, .pdf, .php, .pptx, .py, .rb, .tex, .txt, .css, .js, .sh, .ts( UTF-8 BOM type)
Network block at customer side
Please check from lag from proxy/firewall side. You can take a network trace using developer tools to find 403's or 401's for needed whitelisting of Azure host urls.
Virtual network doc for Azure OpenAI
Could you let us know if the issue still persists with smaller batch upload (with a screenshot please)
Thank you.