Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Feature Request]The server shuts down after one hour of inactivity, which is not good #3428

Open
jiaochenlu opened this issue Jun 18, 2024 · 4 comments
Assignees
Labels
enhancement New feature or request

Comments

@jiaochenlu
Copy link
Contributor

Is your feature request related to a problem? Please describe.
User feedback: The server shuts down after one hour of inactivity, which is not good! Do I have to restart it every morning before our customers wake up?

Describe the solution you'd like
Can be configured not to shut down automatically

@jiaochenlu jiaochenlu added the enhancement New feature or request label Jun 18, 2024
@tyler-suard-parker
Copy link
Contributor

Use case: our chatbot can be used at any time of day by any one of our beta users. If the beta user reports a problem, we want to be able to look at the trace and find out what went wrong. If the tracing server shuts down after one hour of activity, will we lose the trace from any user chats afterward? Will it take 60 seconds for the server to come online, which adds 60 seconds to a chat response?

@YingChen1996
Copy link
Contributor

YingChen1996 commented Jun 27, 2024

Hi, @tyler-suard-parker. Thanks for reporting the issue. Local pfs will auto stop in 1 hour if there's no any request to local pfs (We will remove the behavior). Trace will be stored in database even if the tracing server shuts down, so I think you won't lose the trace. And I think PFS startup will not take 60 seconds

@tyler-suard-parker
Copy link
Contributor

Oh, ok, that makes sense. So when I am using prompflow locally, the local server will auto-stop in one hour if there are no requests to it for promptflow tracing. Those traces recorded by the local server are stored locally.

However, when my web app is deployed, the local tracing server does not matter, because all traces get sent to bot the local tracing server, and my AI Studio tracing. The local server (running on my azure web app) receives the traces and saves them, but they get deleted every time I restart the web app, and I never check them anyways. The promptflow instance running on my web app is also sending traces to my AI Studio workspace, where they are saved. This is why the promptflow tracing server shuts down after an hour, is that correct?

@YingChen1996
Copy link
Contributor

YingChen1996 commented Jul 5, 2024

Hi, @tyler-suard-parker. What you mean is that when you restart the web app, the previous local traces are lost?

BTW, we have removed prompt flow local service (a local web service) auto-stop feature and it won't shut down after an hour without requests in the latest public version. Maybe you can have a try. Thanks

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
3 participants