🏗️ 00 — Azure Well-Architected Framework Overview

🏠 Home

📑 Table of Contents

What Is the Well-Architected Framework?

The Azure Well-Architected Framework (WAF) is a set of quality-driven tenets, architectural decision points, and review tools intended to help solution architects build a strong technical foundation for their workloads on Azure.

WAF is not a checklist to pass — it is a continuous improvement model. It provides guidance across five pillars, each representing a dimension of workload quality. The goal is to help architects make informed tradeoffs that align technology decisions with business objectives.

Official documentation: Azure Well-Architected Framework

The Five Pillars 

%%{init: {"theme":"dark","themeVariables":{"primaryTextColor":"#ffffff","secondaryTextColor":"#e8eaed","nodeBorder":"#27ae60","mainBkg":"#0d2136"}}}%%
flowchart TD
    WAF["🏗️ Azure Well-Architected\nFramework"]
    WAF --> RE["🔒 Reliability"]
    WAF --> SE["🛡️ Security"]
    WAF --> CO["💰 Cost Optimization"]
    WAF --> OE["⚙️ Operational Excellence"]
    WAF --> PE["🚀 Performance Efficiency"]
    style WAF fill:#27ae60,color:#fff,stroke:#1e8449
    style CO fill:#2ecc71,color:#fff,stroke:#27ae60
Pillar Focus Key Question
Reliability Uptime, recovery, redundancy, resiliency at scale Can the workload meet its uptime and recovery targets?
Security Confidentiality, integrity, protection from attacks Is the workload protected against threats and data breaches?
Cost Optimization ROI, spending discipline, rate and usage optimisation Are we maximising the value of our Azure investment?
Operational Excellence Observability, automation, deployment practices, incident response Can we detect and resolve issues in production quickly?
Performance Efficiency Scaling, load testing, meeting demand without waste Can the workload handle changes in demand efficiently?

Each pillar has its own:

  • Design principles — high-level goals that guide architectural decisions
  • Checklist — actionable recommendations (e.g. CO:01–CO:14 for Cost Optimization)
  • Tradeoffs — documented tensions between pillars
  • Design patterns — reusable architectural approaches

How the Pillars Interact

The pillars are interconnected and require deliberate tradeoffs. Optimising for one pillar may create tension with another.

%%{init: {"theme":"dark","themeVariables":{"primaryTextColor":"#ffffff","nodeBorder":"#27ae60","mainBkg":"#0d2136"}}}%%
flowchart LR
    CO["💰 Cost\nOptimization"]
    RE["🔒 Reliability"]
    SE["🛡️ Security"]
    OE["⚙️ Operational\nExcellence"]
    PE["🚀 Performance\nEfficiency"]

    CO <-->|"resiliency vs.\nspending"| RE
    CO <-->|"controls vs.\nlicensing"| SE
    CO <-->|"automation vs.\ntooling cost"| OE
    CO <-->|"capacity vs.\noverprovisioning"| PE

Key tradeoff examples for Cost Optimization:

Tradeoff Description
Cost vs. Reliability Reducing redundancy saves money but increases the risk of downtime. Using budget SKUs may limit achievable SLOs.
Cost vs. Security Removing encryption, simplifying auth, or cutting security tooling lowers spend but weakens defence in depth.
Cost vs. Operational Excellence Cutting testing, documentation, or automation reduces immediate spend but increases bugs, onboarding time, and operational risk.
Cost vs. Performance Under-provisioning or aggressive auto-scale down saves money but may cause latency spikes or capacity failures.

A cost-optimised workload is not a low-cost workload. It is one that maximises return on investment while meeting its functional and nonfunctional requirements.

Workload-Centric Approach

WAF is applied at the workload level — not at the subscription, tenant, or organisation level. A workload is a collection of application resources, data, and infrastructure that together deliver a defined business outcome.

%%{init: {"theme":"dark","themeVariables":{"primaryTextColor":"#ffffff","nodeBorder":"#27ae60","mainBkg":"#0d2136"}}}%%
flowchart TD
    ORG["🏢 Organisation"]
    ORG --> MG["📁 Management Groups"]
    MG --> SUB["📂 Subscriptions"]
    SUB --> RG["📦 Resource Groups"]
    RG --> WL["🎯 Workload\n(WAF applies here)"]
    style WL fill:#2ecc71,color:#fff,stroke:#27ae60

Why workload-level?

  • Different workloads have different business requirements and risk profiles
  • A mission-critical production API has different cost tradeoffs than a dev/test environment
  • Cost decisions must be contextualised to the workload’s SLAs, compliance needs, and business value

The WAF Assessment

Microsoft provides a free, self-service assessment tool — the Azure Well-Architected Review — that helps teams evaluate their workloads against each pillar.

Aspect Detail
URL Azure Well-Architected Review
Scope Per-workload (not per-subscription)
Output Scored recommendations per pillar, prioritised action items
Cost Free
CSA role Facilitate the assessment with the customer, interpret results, and drive follow-up actions

Assessment flow:

%%{init: {"theme":"dark"}}%%
flowchart LR
    A["📝 Select\nworkload scope"] --> B["📋 Answer\npillar questions"]
    B --> C["📊 Review\nscored results"]
    C --> D["🎯 Prioritise\nrecommendations"]
    D --> E["🔄 Implement\n& reassess"]

CSA Tips for WAF Reviews

  • Start with business context — understand the workload’s purpose, SLAs, and business criticality before diving into the pillars
  • Don’t try to cover all pillars in one session — focus on the pillars most relevant to the customer’s current pain points
  • Cost Optimization is often the entry point — many customers come to WAF conversations driven by spending concerns
  • Use the assessment as a conversation starter, not a compliance checklist
  • Document tradeoffs explicitly — when a customer chooses a cheaper option, make sure the reliability or security implications are recorded

WAF and Azure Service Guides

WAF includes service-specific guidance that maps WAF recommendations to individual Azure services. These are useful when a customer asks how to apply WAF principles to a specific resource type.

Service Guide Example Link
Azure Virtual Machines VM service guide
Azure SQL Database SQL DB service guide
Azure Kubernetes Service AKS service guide
All service guides Browse service guides

WAF Workload Perspectives

Beyond the five pillars, WAF also provides guidance for specific workload types that have unique architectural considerations:

Workload Type Focus
AI / ML Discriminative and generative AI models, inference costs, training optimisation
SaaS Multi-tenancy, billing models, ISV cost structures
Mission-Critical Always-available workloads where cost tradeoffs are bounded by strict SLOs
SAP on Azure SAP-specific sizing, licensing, and migration cost considerations
HPC Burst compute, Spot VM strategies, high-density scheduling
Azure VMware Solution Hybrid migration staging, reserved node pricing

Where Cost Optimization Fits

Cost Optimization is one of five pillars, but it often serves as the primary driver for customer engagement with WAF. Customers rarely approach a CSA saying they want to improve operational excellence — they say they want to reduce their Azure bill.

This creates a natural entry point:

  1. Start with Cost Optimization — address the customer’s immediate concern
  2. Surface tradeoffs — show how cost decisions impact reliability, security, and performance
  3. Broaden the conversation — use the cost review to introduce the full WAF framework
  4. Establish continuous improvement — position cost optimisation as an ongoing discipline, not a one-time exercise

The next section dives deep into the Cost Optimization pillar — its design principles, checklist, tradeoffs, and maturity model.

Next → 01 — Cost Optimization Deep Dive

🏠 Home