Vigith Maurice

Why do we need a prometheus server in Numaflow? Can't we assume/mandate that a metrics provider that supports PromQL is provided?

thank you, @veds-g for driving it

> Hey! I’m currently exploring gRPC as part of a project involving llama.cpp, and this issue looks like a great way to understand it better. If no one is working...

> Indeed very hard to tell from "test" labels whether is related to numaflow core tests (go/sdk), or test framework for users (pipelines/udf). We usually tag `area/sdk` when it is...

Moved the issue to Python SDK repo. Updated the summary. Please let me know if you want me to add more details.

the moment we add handler to struct, `Clone` will no longer be applicable

This processing rate is coming from the daemon server based on how we track data movement based on reads. We might have to somehow distinguish not processing vs not reading....

This involves https://github.com/PyO3/pyo3/. Are you familiar with it?

Actually it is not completed, I just left it open as a reminder for us to improve the design and implementation of Sink retry.

Resolved via #3021