There is seriously no one else experiencing API errors?

I have an App that makes 15 calls in a row that never would have a bug of just shitting out before devday. now after devday, it will just stop by the 10th call and never load the rest of the calls for some reason. I mean I can’t even use my software that was working perfectly now because it won’t make all the calls

I think it has to do with the overload of traffic after the DevDay if you don’t get any specific errors. Also, a few days ago, OpenAI noticed that they were experiencing strange activity patterns that looked like a DDOS attack. So, that might also play a role in why you have problems. You can clearly see from the screenshot below that the uptime got worse after the DevDay. You can check OpenAI status here.

Hi there

I understand that and I frequently check the above link to check for server updates. The problem is the servers are reporting “All Systems Operational”, and that is not the case for me. If you have 15 different calls from a for loop which I do - it will just stop by the 10th or 12th, no errors, it just stops for like…2 minutes, and then continues. This was not a problem before devday and now is all of a sudden like a major problem I’m running the program and it only succeeds like 30% of the time as of yesterday.

It’s just suspicious to me but…if no one else is having any issues at all then i don’t know

Thing to check at the moment is that you are below your rate limit errors, and possibly use the stream=true option.

When streaming the results in you are able to check for failed replies and issue a retry rapidly, say a 5 second wait for tokens. That way you can manage any failed calls.

If you have not already, you may need to update your openai library installation to the latest version to ensure all of the new error checking is being performed.