Description
Note: if you have a question or need support please post in the discussions area.
Describe the bug
When accessing the application through cloudflare, the GetOperationEvents request will timeout and will end up causing the live updates of the application to not function.
It seems that GetOperationEvents request is a long running request which seems to constantly send data, so when accessing locally without cloudflare, the live updates of the application work as intended with this request not timing out but still finishing after a few minutes. However, with cloudflares timeout limits in place, this request always fails leading to application live update issues.
To Reproduce
Steps to reproduce the behavior:
- Access the application through cloudflare (e.g. public hostname on cloudflared tunnel)
- Observe that after 1.7 minutes, the request for GetOperationEvents timeouts and a new one is created
- Observe that when clicking on "Prune Now" or other buttons, a refresh of the page is required for the prune to show in the page and its status.
Expected behavior
When accessing the web page through cloudflare, all live updates should work as expected and the application is responsive.
Platform Info
- OS and Architecture Unraid 7.0.1 in Docker container
- Backrest Version: 1.8.0
Additional context
It seems that as this is a long running request, it could be replaced with a websocket which will be allowed through cloudflare with no issues and would allow for instant communications of events and live updates without ever timing out.