We have a continuous issue with our deployment of phoenix. After a while users are presented with the following screen, they then have to go back to the login page and log back in - this is a real pain. We don't have any errors in our logs and are really struggling to debug what could be causing this - any suggestions where to start?
This issue is likely caused by the database connection (e.g., PostgreSQL) timing out after a period of inactivity, which results in the Phoenix application not properly handling reconnection when user interaction resumes. Users then see an error and must refresh or log in again. The recommended first step is to check your database connection settings (timeouts, pool size) and ensure Phoenix is configured to handle idle connections gracefully. Refreshing the page typically resolves the error, confirming it's a session/connection issue rather than an application bug. See the detailed discussion and error trace in this GitHub issue.Would you like more detail on how to adjust your database or Phoenix settings to mitigate this problem?Sources: - bug UI Error After Period of Inactivity
馃挕聽Hint: Mention RunLLM in the thread for followups.
we are hosting this ourself
let me check the db setup
we are using the SQLite setup - would moving to PG help with this?
Definitely recommend using PG if you are hosting. SQLite is a single file so it's ideal for running on your computer or in a notebook but not very good for hosting. It's quite a bit faster and performant.
:ty:
