Su Myat Hlaing Greetings!
I understand your concern.
ECONNRESET errors indicate that the connection was forcibly closed by the server. This could happen if there is a network instability or server-side restrictions on connection duration.
Did you try implementing a retry logic to handle transient network issues?
Is there a default time limit or timeout setting in the API that could be impacting the upload of files?
AFAIK, Azure OpenAI service has a maximum file size limit of 512MB per file for uploads and 200 MB via Azure AI Foundry portal. However, there isn't a specific timeout limit mentioned in file uploads. The timeout error you're encountering might be due to network latency or other factors affecting the upload process.
Why am I seeing a difference between the Playground (UI) upload success rate and the API behavior?
This depends on your network. The UI might manage requests differently, such as by batching them that reduces the likelihood of hitting server-side limits or timeouts.
Are there additional file size or time restrictions for the Azure OpenAI API when uploading files for Assistant File Search?
You can find more details for applicable limits on Azure OpenAI Service quotas and limits
Is the error related to DNS (getaddrinfo ENOTFOUND), and how can I resolve this issue in my API requests?
To isolate the issue, try using a different DNS server or IP address to see if that resolves the issue.
If that doesn't help, For a deeper investigation and immediate assistance on this issue, please file a support request @ https://aka.ms/azsupt?