Björn Rabenstein

Results 1096 comments of Björn Rabenstein

First of all, thanks for the high-quality PR with all the test cases, comprehensive documentation, and thoughtful explanations. I have a bunch of considerations at various levels of detail. Let's...

After a few days of letting my thoughts settle, I'll try to give this at least some preliminary conclusion. First of all, I will add this as an agenda item...

Yeah, that's kind of what troubles me. I think for rate calculation, it's deliberate to not take staleness markers into account. As said above, that's the beauty of counters. We...

As discussed with @MichaHoffmann , Monte-Carlo integration works with a random distribution of samples, representative for the full time series. However, what we are trying to solve here is precisely...

> @beorn7 could you summarise where all the deliberations reached? Is this addition acceptable? If I had to decide this now, and if I had to decide alone, I would...

Almost a year after adding this as an agenda item for the dev-summit, it was finally discussed. See [notes](https://docs.google.com/document/d/1uurQCi5iVufhYHGlBZ8mJMK_freDFKPG0iYBQqJ9fvA/edit?pli=1&tab=t.0#bookmark=kix.wv2nhg4q6k1t). Sadly, the outcome I had hoped for didn't happen. Ideally, one...

Hello from the bug scrub! To debug this, we need exactly the data that got scraped. Could you `curl` that data and paste it here? Note that the out-of-order feature...

There is a high bar for introducing "yet another modifier". I think at this low level, it is easier to solve this kind of alignment by allowing arithmetic for durations...

I mostly agree with your reasoning. I just thing we have to build the building from the ground up and not start at the top floor. The basement is essentially...

Note that arithmetic on durations work now, with the `--enable-feature=promql-duration-expr` flag set, see https://prometheus.io/docs/prometheus/latest/feature_flags/#promql-arithmetic-expressions-in-time-durations However, arithmetic is not (yet) supported for the argument of the `@` modifier, and you cannot...