Jon Gjengset

The issue with giving read-your-own-writes currently is actually that views don't have a well-defined time for you to query. Even if we gave an ID back with every write, that...

We don't actually! Or rather, Noria propagates all effects of writes throughout all views, but it doesn't know _which_ write any given update corresponds to. A view learns, for example,...

@ms705 there sadly isn't really a way to do this in a good way. Maybe the closest approximation is using [`serial_test`](https://docs.rs/serial_test_derive/0.2.0/serial_test_derive/index.html). There's https://github.com/rust-lang/rust/issues/43155, but I don't know that anything has...

Hi @bartsmykla! Sure thing, feel free! I'm having second thoughts about labeling this as "good first issue" though, as figuring out what logging levels are appropriate in each instance, and...

Also: - Read retries + time-to-completion - Domain frequency + processing time per destination node

@jbcden Thanks for the offer! Thinking some more about this, I suspect this will actually require some relatively large-scale system refactoring to allow capturing all the metrics we care about....

@ekmartin has been thinking about this I think.

Huh, that is _very_ weird indeed. What platform are you on? Can you try adding `-v` to where you launch `noria-server`? That should give you some more details about what...

Also, can you try the `griddle` branch and see if the same issue occurs there?

Ah, sorry, yes, you'll also want to swap the `#` on these two lines to enable trace logging: https://github.com/mit-pdos/noria/blob/f54dfcef4057b490ccb43f4a31c6aaf35c923fb7/server/Cargo.toml#L42-L41