jrhee17

turns out we're already leaving metrics 😅 closing https://github.com/line/centraldogma/blob/aa4b789f2714ec0742aa94d8c1709615e1b0e4f2/server/src/main/java/com/linecorp/centraldogma/server/CentralDogma.java#L795

@ikhoon also found another executor which isn't being monitored - it would be nice to also add metrics https://github.com/line/centraldogma/blob/e8535d41c242573de45645bb933f84011eec0ebc/server/src/main/java/com/linecorp/centraldogma/server/internal/mirror/DefaultMirroringService.java#L109-L121

I believe this PR is (finally) ready for review. Please take a look at your convenience 🙏

Let me close this PR and create a separate one 🙇

Looks good so far 👍 I just wanted to confirm my understanding on the concept of `maxConcurrency`. I understood `maxConcurrency` to be the maximum number of child `StreamMessage`s a `FlatMapStreamMessage`...

> The total number of ongoing requests should be equal to maxConcurrency - buffer.size() I see. I think this is what was confusing me 😅 Assuming an initial state where...

> In that case there would be a request for each subscriber. At the start the buffer is empty, so you would get maxConcurrency - buffer.size() -> 3 - 0...

Can you also check the CI failures? I think this PR should be almost done once my comments are addressed 🙇

Let me leave some additional information for those interested. The goal of this pull request is to add support for `@Decorator` annotation for thrift services. Assume that we define a...

Hi, I think this is probably better to ask to the opentelemetry team. When I tried setting the exporter to `logging` (i.e. `-Dotel.traces.exporter=logging`) I found that actually armeria server isn't...