Paschalis Tsilias

The underlying issue in loki was solved via grafana/loki#7771 which is included in the updated version we depend on from #2590. Closing as completed.

I'm closing this issue as completed; somewhere along the line we've fixed this issue. A config like the following will only report a single error ``` local.file "secret_key" { filename...

I suppose one of the most basic ideas is to strategically place some benchmarks, and compare them between commits, right? There are some built-in tools like `benchstat` and `benchcmp` which...

Hmm, you're right. The `benchstat` tool actually reports differences only if they're statistically significant which should somewhat normalize this fluctuation. To do so it requires 4-5 runs per benchmark *before*...

So do we think this discussion should be moved to the spec first then? Just for posterity, there is the possibility of using a [userinfo](https://datatracker.ietf.org/doc/html/rfc3986#section-3.2.1) to pass in auth (like...

Yeah, agreed. I'll move the discussion over to the spec repo as a new issue, it's gonna be my first one there. Thanks! 🙏 **Edit:** opened https://github.com/open-telemetry/opamp-spec/issues/194

I couldn't reproduce any failures even after >40k tries with `$ go test -run TestAgentIdentification/ws -race -count 10000 ./...`. Also, these old action runs no longer have any logs associated...

Closing this issue as completed as per the original author's comment.

It looks OK to add this since the measurement does include the type in the matching faro.receiver component. https://github.com/grafana/agent/blob/287b1c68cae77768d70ff20f7d753db943586393/component/faro/receiver/internal/payload/payload.go#L228 @vanugrah Could you please also add a CHANGELOG entry under Enhancements...

I'd also +1 trying to wire this in instead.