[GH-ISSUE #1376] feat: backend timeout error handling #99461

Closed
opened 2026-05-17 21:28:27 -05:00 by GiteaMirror · 3 comments
Owner

Originally created by @shaul75 on GitHub (Apr 1, 2024).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/1376

Description

Whenever the API request to the backend takes a while to return any data (usually because the model wasn't loaded yet), and there's a "504 Gateway Time-out" error, the client does not handle it and the user would see the loading skeleton indefinitely, with no indication of a problem unless you open the dev tools.

I suggest adding error handling that'll show a failure message and a retry button to make it clear the request to the backend failed.

Originally created by @shaul75 on GitHub (Apr 1, 2024). Original GitHub issue: https://github.com/open-webui/open-webui/issues/1376 ## Description Whenever the API request to the backend takes a while to return any data (usually because the model wasn't loaded yet), and there's a "504 Gateway Time-out" error, the client does not handle it and the user would see the loading skeleton indefinitely, with no indication of a problem unless you open the dev tools. I suggest adding error handling that'll show a failure message and a retry button to make it clear the request to the backend failed.
Author
Owner

@tjbck commented on GitHub (Apr 1, 2024):

Steps to reproduce would be extremely helpful here!

<!-- gh-comment-id:2029357390 --> @tjbck commented on GitHub (Apr 1, 2024): Steps to reproduce would be extremely helpful here!
Author
Owner

@shaul75 commented on GitHub (Apr 1, 2024):

Steps to reproduce would be extremely helpful here!

I don't know of an easy way to force the backend to return 504 there, for me it happens pretty consistently when loading large LLMs (>20GB) during the first chat message.

I can look into it and create a PR myself when I have some free time (probably later today).
I've seen other response codes are already handled in the UI so it should be a small fix

<!-- gh-comment-id:2029418459 --> @shaul75 commented on GitHub (Apr 1, 2024): > Steps to reproduce would be extremely helpful here! I don't know of an easy way to force the backend to return 504 there, for me it happens pretty consistently when loading large LLMs (>20GB) during the first chat message. I can look into it and create a PR myself when I have some free time (probably later today). I've seen other response codes are already handled in the UI so it should be a small fix
Author
Owner

@tjbck commented on GitHub (Apr 19, 2024):

Closing in favour of #1611

<!-- gh-comment-id:2066389320 --> @tjbck commented on GitHub (Apr 19, 2024): Closing in favour of #1611
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#99461