3.x Add note to MP performance (and SE?) about using thread pool metrics for monitoring

[tjquinno]

Environment Details

• Helidon Version: 3.x
• Helidon SE or Helidon MP
• JDK version:
• OS:
• Docker version (if applicable):

---

Problem Description

The MP performance tuning doc page mentions customizing thread pool settings and turning on FINE-level logging so the log output reports the non-default settings.

A good addition to that section would be a brief description of the Helidon-provided thread pool metrics. The ExecutorServiceMetricsObserver class in helidon/metrics/metrics is where metrics are exposed (based on the corresponding methods the Java ThreadPoolExecutor type exposes). The metric naming is straightforward but the tags are a little interesting and will need to be explained because of the need to describe multiple supplier categories, suppliers within them, and pools within suppliers so we'd need to describe that a bit.

• active-count
• completed-task-count
• pool-size
• largest-pool-size
• task-count
• queue
  • remaining-capacity
  • size

Steps to reproduce

Visit the linked page above.