Andrew Wilkins

Sorry for the delay, and thanks for raising the issue. This is more an issue with the Logs app in Kibana. I believe it's applying the "generic webserver" rules to...

@AdamKorcz thanks for opening the issue! The fuzzing program I wrote a long time felt into disarray precisely because we haven't been running it continuously. I'm not _entirely_ convinced it's...

> I realize that for cases where the URL doesn't include path parameters, the getting started experience is nicer when using the path as the transaction name. However, the risks...

There's a couple of thorny issues here to work through: - there's no in-process context propagation in the sarama API - we don't yet have precedent or design on how...

I've done a little more investigation of tracing for sarama. Here are my current thoughts: - We will provide functions for injecting trace headers into, and extract from, messages. Kafka...

> As the unit for that setting is bytes, not the event count, and it's already set to a relatively modest value, I don't api_request_size needs to be called out...

Applications might search time-based (foo-2021-05-04) or ILM indices (foo-00001), which won't have a stable name that's useful in aggregations. I'm not sure if `GET /customers/_doc/42 -> GET /customers/_doc` is going...

Still relevant. Thanks.

The Go agent currently has no config for this, it just grabs the whole stack track. https://github.com/elastic/apm-agent-go/issues/526 is open to implement the config. For Go, including `truncated: true` would be...

Thanks for opening the issue! I agree, this would be nice to have. Right now the protocol between agents and server does not support sending arbitrary logs, which is why...