Add Windows disk I/O metrics to ResourceMonitoring

[makazeu]

Closes#5842

Add the below new Windows metrics to the ResourceMonitoring library:

• system.disk.io
• system.disk.operations

Will try to support below metrics in the future:

• system.disk.io_time
• system.disk.operation_time
• system.disk.pending_operations

Also, will support Linux in the future

Tested in Grafana + Prometheus with OpenTelemetry.Exporter.Prometheus.HttpListener

Tested in Geneva with OpenTelemetry.Exporter.Geneva

Microsoft Reviewers: Open in CodeFlow

[makazeu]

cc @evgenyfedorov2

[RussKie]

Primarily code style related comments.

[evgenyfedorov2]

Hi @makazeu , thanks for pickup up this task!

I am busy with other things right now, could you please don't merge this until I have a chance to review it? Will try to review this or next week.

Thanks!

[makazeu]

Hi @makazeu , thanks for pickup up this task!

I am busy with other things right now, could you please don't merge this until I have a chance to review it? Will try to review this or next week.

Thanks!

Sure, Evgeny. This is still a draft yet, I will not merge it until it has more unit tests and you have reviewed it.

[makazeu]

This PR is now ready for review!

Edit: Add screenshosts for my tests with Grafana + Prometeus, and Geneva.

[makazeu]

Hi @evgenyfedorov2 If you get a chance, could you please help review this PR? I'd really appreciate it!

[evgenyfedorov2]

Why are screenshots from tests different in absolute values? It is max 50M in Geneva, but only about 3.5M in Grafana.

[makazeu]

Why are screenshots from tests different in absolute values? It is max 50M in Geneva, but only about 3.5M in Grafana.

Because those screenshots were taken at different times. I remember the Geneva screenshot was added a day later.

[evgenyfedorov2]

I think the code itself looks good, can we just have screenshots taken at the same time, please?

[makazeu]

I think the code itself looks good, can we just have screenshots taken at the same time, please?

Hi @evgenyfedorov2, I updated the screenshots in the PR description, now they were taken at same time. In the two charts, their trends are the same, but due to differences in the aggregation precision, they vary slightly in details. I can't make them completely identical, but overall, they are quite similar.