Steve Gordon

> 1. Add a dotnet.cpu.time metric to provide CPU usage info: [Add initial experimental .NET CLR runtime metrics #1035 (comment)](https://github.com/open-telemetry/semantic-conventions/pull/1035#discussion_r1634044798) Yep, I still need to add `dotnet.cpu.time` and `dotnet.cpu.count`. The...

No problem, @lmolkova. Thanks again for the feedback. I've updated per your suggestions and also snuck in the new CPU metrics.

@lmolkova I'm thinking `dotnet.thread_pool.queue.length` unit should be `{work_item}` given what the queue length represents. Would you concur?

I've opened a PR for the implementation at https://github.com/dotnet/runtime/pull/104680. It's a draft for now while we finalise the last naming discussion. We'd need to get this approved quite soon to...

Thanks, @lmolkova. I've pushed up all the suggested changes.

Given that we have an implementation ready to go, should we consider adding these as `stable`?

I'd be interested in starting work on adding these.

@pyohannes If I've followed the Java story around this, I believe their instrumentation defined many of the metrics later added to the semantic conventions, and I saw [a comment](https://github.com/open-telemetry/semantic-conventions/issues/535#issuecomment-1815166292) from...

My thoughts were to first review what existing metrics are available (not many as most are event counters I believe), which were produced from contrib instrumentation and also to compare...

@pyohannes Yep, I agree. Unfortunately, the timing of those SIGs is very difficult for me to attend but I'll try to get to one soon and put this on the...