Assistant timeout or bad request

Question

I had issues working on the assistant on the app we are developing so I went to check on the Azure AI Foundry if it was a mistake on my behalf.

The platform either goes in timeout or gives the following error:

**Failed to create message in the thread** unknown_parameter: Unknown parameter: 'content'. **Trace ID :** bf1f162f-e664-449b-bc58-244479bcf88b**Client request ID :** 174d59bc-4043-4240-aafe-7e3f8170178d**APIM request ID :** 16a5cb1f-17f7-42b3-9cc8-40937bd83aab

Stack trace:

Error: Request failed with status code 400 at ht (https://ai.azure.com/assets/manualChunk_data-fetch-bd489cf3.js:14:85613) at dt (https://ai.azure.com/assets/manualChunk_data-fetch-bd489cf3.js:14:85796) at XMLHttpRequest.S (https://ai.azure.com/assets/manualChunk_data-fetch-bd489cf3.js:15:1655)

I tried with both gpt-4o-mini v(2024-07-18) and gpt-4o v(2024-05-13) and they behave the same way.
We work on the EUW server (France). On the other resource group in swc it seems like there is no problem.

Answer 1

Hi Jacopo Fichera

It seems you are experiencing issues with timeouts and errors while using the Azure AI Foundry with the assistant models. If you exceed the index quota for your service, you may encounter errors. You can either delete unused indexes or upgrade your service for higher limits. if the input documents are too large, it might lead to timeouts. Consider breaking down your input documents into smaller sizes. Since you mentioned that the issue occurs on the EUW server (France) but not on another resource group, it might be worth checking if there are specific issue with prompts or network latency on that group

Kindly refer below link: troubleshooting

Thank You.