Skip to content

Design

The cross-cutting architectural foundation that governs both delivery tracks.

This section is track-agnostic. Everything here applies equally to the SCOM Management Pack and the Azure Monitor Health Model — the two tracks are different implementations of the same conceptual design.

If you're new to the project, read this section first, in order:

  1. Scope & Topology — what we monitor and at what granularity. The 3-layer, ~25-entity infrastructure model. Workloads are explicitly out of scope.
  2. Health Model — how health is structured (dimensions, states, rollup policy, impact, suppression).
  3. Signal Catalog — the master list of every signal that drives a health state, with source, threshold, and parity across both tracks.
  4. Customization — how operators tune thresholds without forking the product. Sealed MP + override pack tiers (SCOM); Bicep params + tier files (Azure Monitor).
  5. Concept Mapping (SCOM ↔ AzMon) — side-by-side translation of every architectural concept across the two tracks. The dual-track Rosetta Stone.
  6. Decisions (ADRs) — the ten Architecture Decision Records that lock down each of the choices above.

Why a separate Design section?

Three failure modes this section prevents:

  • Track drift — without a single source of truth for "what does healthy mean", the SCOM and Azure Monitor implementations would diverge over time. Every threshold and signal in this section has a single canonical name and value, mapped 1:1 across both tracks.
  • Buried decisions — ADRs that live only in a decisions/ folder on disk are invisible to readers who only see the rendered docs. Putting them in the navigation makes the reasoning behind every architectural choice discoverable.
  • Customization as a bolt-on — treating customization as a documentation afterthought produces forked deployments. By making it a first-class design topic with cross-track parity guaranteed by ADR 0007 and ADR 0008, customization becomes a product feature.

Design principles

The choices in this section are guided by five principles:

  1. One model, two tracks — the SCOM MP and Azure Monitor Health Model are different surfaces over the same underlying entity graph, signal catalog, and rollup policy.
  2. Infrastructure only — workloads (VMs, AKS pods, applications) are deferred to companion MPs. See ADR 0001.
  3. Customization without forking — every threshold and behavior is parameterized. Customers ship overrides, not rebuilt MPs.
  4. Cross-track parity is a hard requirementVolume.FreeSpace.WarnPercent (SCOM) and volumeFreeSpaceWarningThresholdPct (Azure Monitor) are guaranteed to mean the same thing. See ADR 0007.
  5. Cite first-party sources — every signal, threshold, and prerequisite is traced back to Microsoft Learn, Azure Local product docs, or a named upstream reference. The REFERENCES file is the bibliography.

Status

Layer Phase 2 status
Scope & topology (ADR 0001) Proposed — locks 3 layers, ~25 entities, infra-only
Health model (ADR 0003) Proposed — worst-state rollup with documented exceptions
Signal catalog Drafted — ~60 signals across 3 layers
Customization (ADR 0008) Proposed — three tiers (Lab / Standard / Strict)
Cross-track parity (ADR 0007) Proposed — naming convention locked
Decisions 0002–0010 Proposed — see Decisions index

When all ten ADRs move from Proposed to Accepted the project enters Phase 3 (SCOM authoring) and Phase 4 (Azure Monitor authoring) in parallel.