Processing large RAG documents fails due to Cloudflare timeout

[Cyberes]

What happened?

Cloudflare enforces a 100 second timeout for connections through their network and if a server does not respond within 100 seconds the connection will be closed. But processing documents for RAG may take longer than that, meaning these documents will always fail to be processed because Cloudflare will give a 524 (timeout) error.

The solution to this would be to have the server send status updates so that the connection stays open. I would do SSE where the server sends things like "processing" every 10 seconds then ends with a "completed".

Version Information

ghcr.io/danny-avila/librechat-dev latest 2fa722380692 11 days ago 1.56GB
ghcr.io/danny-avila/librechat-rag-api-dev-lite latest 2f812738600a 2 weeks ago 1.67GB
ghcr.io/danny-avila/librechat-dev e05ac102518f 3 weeks ago 1.17GB
ghcr.io/danny-avila/librechat-dev 69836bf6d276 3 months ago 1.04GB
ghcr.io/danny-avila/librechat-dev 7ee5765fe941 3 months ago 1.04GB
ghcr.io/danny-avila/librechat-dev 7b572ea1536a 4 months ago 919MB

Steps to Reproduce

1. Create 2 MB text file with lots of lines
2. Upload to LC through CF

What browsers are you seeing the problem on?

Chrome

Relevant log output

upload error

Screenshots

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[danny-avila]

Thanks for the suggestion. I've recategorized this as a feature request since Cloudflare's 100-second timeout is a documented infrastructure constraint rather than a bug. While SSE could stream processing updates after a file upload, it's not a usual pattern for file uploads and may not address the core issue for long processing times.

An alternative would also be to implement async processing patterns (immediate response with job ID, then status polling) to handle long-running RAG operations within these timeout constraints.