docs: docs for machine id metrics and hw-health metrics fetch

[yoks]

Documentation updates for #2433 and #2168

Related issues

#2433 #2168

Type of Change

• Add - New feature or capability
• Change - Changes in existing functionality
• Fix - Bug fixes
• Remove - Removed features or deprecated functionality
• Internal - Internal changes (refactoring, tests, docs, etc.)

Breaking Changes

• This PR contains breaking changes

Testing

• Unit tests added/updated
• Integration tests added/updated
• Manual testing performed
• No testing required (docs, internal refactor, etc.)

Additional Notes

[coderabbitai]

Walkthrough

Expanded monitoring documentation with hardware entity metrics configuration and output details, plus opt-in per-object health metrics behavior and lifecycle semantics.

Changes

Monitoring health documentation

Layer / File(s)	Summary
Hardware entity metrics documentation docs/operations/monitoring-health.md	Adds entity discovery and collector defaults, Redfish *Metrics collection behavior, configuration, Prometheus naming and labels, OTLP correspondence, and first-discovery requirements.
Per-object health metrics documentation docs/operations/monitoring-health.md	Documents classification-based unhealthy-object metrics, configuration, labels, default behavior, and series disappearance.

Estimated code review effort: 2 (Simple) | ~10 minutes

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title is concise and clearly references the documentation updates for machine ID and hardware-health metrics.
Description check	✅ Passed	The description is directly related to the documented metrics and linked upstream PRs.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

🔍 Container Scan Summary

No Grype artifacts were found to aggregate.

[github-actions]

🌿 Preview your docs: https://nvidia-preview-pull-request-3377.docs.buildwithfern.com/infra-controller

[coderabbitai]

Actionable comments posted: 4

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@docs/operations/monitoring-health.md`:
- Around line 595-597: The monitoring documentation inaccurately claims
classification-based emission keeps cardinality bounded. In the opt-in emission
description, replace “bounded” with wording that it helps contain cardinality,
clarify that object_id still scales series with fleet size, and note that
multiple enabled classifications multiply per-object series.
- Around line 610-617: The rack alert example is incompatible with the
documented metric labels because it neither filters to machines nor identifies
racks. Update the example prose and PromQL to describe a fleet-wide
hardware-unhealthy machine count, or add/document a `rack_id` label and group
the query by rack so it can correctly alert on per-rack totals.
- Around line 227-232: Clarify the sink-specific query contract in the section
describing series labels and OTLP. State that Prometheus exposes inventory
context as metric labels, while OTLP exports it as resource attributes; document
the exact attribute-to-label mapping and query differences, or limit the claim
to shared metric names without asserting identical Grafana queries.
- Around line 604-608: The monitoring documentation’s eviction timing statement
is inaccurate: update the paragraph near the per-object classification metrics
to state that series are retained for the registry’s hold period, which is
configured slightly longer than the state controllers’ `metric_hold_time`,
rather than expiring exactly at that value. Preserve the existing behavior for
empty classifications and aggregate metrics.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 8e57c192-97ff-41ed-a3de-ae7a282c8c64

📥 Commits

Reviewing files that changed from the base of the PR and between 1a1b1e2 and 08c2431.

📒 Files selected for processing (1)

• docs/operations/monitoring-health.md

[polarweasel]

Couple quick little things, then good to go

[coderabbitai]

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

docs/operations/monitoring-health.md (1)

616-620: 🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win

Do not describe in_use="false" as allocation blocking.

The documented in_use label only indicates whether a tenant instance uses the machine; it does not establish allocation eligibility. This query lists unused machines carrying the Hardware classification, not necessarily hosts blocked from allocation. Narrow the prose accordingly, or query the classification that explicitly represents allocation prevention if that is the intended contract.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@docs/operations/monitoring-health.md` around lines 616 - 620, Update the
prose introducing the PromQL query to describe it only as listing unused
machines with the Hardware classification, not as hosts blocked from allocation.
Keep the query unchanged unless an explicit allocation-prevention classification
is available and intended by the documentation contract.

Source: Path instructions

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Outside diff comments:
In `@docs/operations/monitoring-health.md`:
- Around line 616-620: Update the prose introducing the PromQL query to describe
it only as listing unused machines with the Hardware classification, not as
hosts blocked from allocation. Keep the query unchanged unless an explicit
allocation-prevention classification is available and intended by the
documentation contract.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: f7dd249b-b88b-46a0-9ca7-8e075a4c8e57

📥 Commits

Reviewing files that changed from the base of the PR and between 0937c8f and 5f130f0.

📒 Files selected for processing (1)

• docs/operations/monitoring-health.md