Skip to content

Stabilize Workloads

Stabilize is about getting to Azure quickly and safely. The applications do not change. The databases stay compatible. But the infrastructure underneath becomes modern, managed, and cost-optimized. In the underlying Horizons model, this path maps to Horizon 1.

What Moves

On-PremisesAzure TargetMigration Tool
Windows Server VMsAzure Virtual MachinesAzure Migrate
SQL Server databasesAzure SQL Managed InstanceMI Link / Azure DMS
Network configurationAzure Virtual Network + NSGsAzure Migrate

The Architecture

%%{init: {'theme':'neutral'}}%%
graph TB
  classDef azure fill:#0078d4,stroke:#005a9e,color:#fff
  classDef db    fill:#038387,stroke:#025356,color:#fff
  classDef ops   fill:#deecf9,stroke:#0078d4,color:#003459
  subgraph vnet["Azure Virtual Network"]
    AVM(["Azure VMs<br/>.NET Framework Apps"]):::azure
    SQLMI[("Azure SQL<br/>Managed Instance")]:::db
  end
  subgraph plat["Platform Services"]
    MON["Azure Monitor"]:::ops
    BAK["Azure Backup"]:::ops
  end
  AVM -->|"App traffic"| SQLMI
  AVM -.->|"Telemetry"| MON
  SQLMI -.->|"Telemetry"| MON
  SQLMI -.->|"Backup"| BAK
  style vnet fill:#e6f3ff,stroke:#0078d4
  style plat fill:#f0f9f0,stroke:#107c10

Why SQL Managed Instance

SQL Managed Instance is the key to the Stabilize path. It provides near-100% compatibility with on-premises SQL Server — which means databases move without application code changes.

What you keep:

  • SQL Agent jobs
  • Cross-database queries
  • CLR assemblies
  • Linked servers (within the managed instance)
  • Database mail

What you gain:

  • Automated patching and backups
  • Built-in high availability
  • Transparent data encryption by default
  • Elastic scaling without downtime
  • Pay-per-use instead of upfront licensing

Quick Wins

The Stabilize path delivers measurable value within weeks:

  • Right-sizing: VMs are provisioned to match actual utilization, not peak-plus-headroom estimates from five years ago
  • Reserved instances: Commit to 1-year or 3-year terms for eligible, predictable workloads after sizing is validated
  • Automated operations: Patching, backups, and monitoring move from manual processes to managed services
  • Decommission on-premises: Reduce data center footprint and associated facilities costs

Before and After

DimensionBefore (On-Premises)After (Stabilize)
InfrastructureSelf-managed VMs, manual patchingAzure VMs with automated patching
DatabaseSQL Server (self-managed)SQL Managed Instance (fully managed)
BackupsManual, on-siteAutomated, geo-redundant
MonitoringBasic or manualAzure Monitor with alerts
Cost modelCapEx, over-provisionedOpEx, right-sized
AnalyticsManual spreadsheet exportsSQL MI Mirroring → Fabric
Typical timeline4–8 weeks

What Comes Next

Once workloads are running in Azure, you have two paths:

  1. Stay in Stabilize — For stable, low-change workloads, this is the right answer. Add Fabric integration via SQL MI Mirroring to unlock analytics without any further changes.
  2. Evolve to Transform — For workloads that need elasticity, modern DevOps, or cloud-native capabilities, plan a Transform modernization.

← Back to Modernization Paths · Next: Stabilize + Fabric →