Backup & Disaster Recovery

The SMB Ready Foundation prioritizes cost optimization over high availability. Recovery involves rebuilding from Infrastructure as Code rather than automatic failover.

Recovery Objectives

Metric	Target	Notes
RPO	24 hours	Daily VM backups
RTO	2–4 hours	IaC redeployment
Availability	Best effort	No SLA commitment

RTO by Tier

Tier	RTO Target	Services
Critical	2 hours	VPN connectivity (if deployed)
Important	4 hours	Azure Firewall, network routing
Standard	8 hours	Customer VMs, workloads

VM Backup

VMs tagged with Backup: true are automatically enrolled via Azure Policy (smb-backup-02, DeployIfNotExists).

Setting	Configuration
Backup Vault	rsv-smbrf-smb-{region}
Policy	DefaultVMPolicy
Schedule	Daily @ 02:00 UTC
Instant Recovery	2 days
Daily Retention	30 days
Weekly Retention	12 weeks (Sunday)
Monthly Retention	12 months (1st of month)
Geo-Redundancy	LRS (Locally Redundant)

Restore a VM

# Set vault context
$vault = Get-AzRecoveryServicesVault -Name "rsv-smbrf-smb-swc" `
  -ResourceGroupName "rg-backup-smb-swc"
Set-AzRecoveryServicesVaultContext -Vault $vault

# Find recovery points
$container = Get-AzRecoveryServicesBackupContainer `
  -ContainerType AzureVM -FriendlyName "vm-name"
$item = Get-AzRecoveryServicesBackupItem `
  -Container $container -WorkloadType AzureVM
$rp = Get-AzRecoveryServicesBackupRecoveryPoint -Item $item

# Restore
Restore-AzRecoveryServicesBackupItem -RecoveryPoint $rp[0] `
  -TargetResourceGroupName "rg-spoke-prod-swc" `
  -StorageAccountName "strecoveryswc" `
  -StorageAccountResourceGroupName "rg-backup-smb-swc"

Infrastructure Recovery

Infrastructure is defined as code and stored in Git — no backup needed:

Component	Recovery Method
Bicep/Terraform templates	azd up redeploy from Git
Azure Policies	Auto-applied on deployment
Network config	Auto-applied on deployment
Firewall rules	Defined in IaC templates

Full infrastructure recovery:

cd infra/bicep/smb-ready-foundation   # or terraform
azd env set SCENARIO full
azd up

Note

Infrastructure redeployment takes 5–50 minutes depending on the scenario. This is the “rebuild-from-IaC” DR strategy documented in ADR-0001.

Data Not Backed Up

Resource	Reason	Mitigation
Log Analytics data	Ephemeral operational data	Export critical queries to saved searches
Key Vault secrets	Soft delete + purge protection enabled	90-day recovery window
Azure Migrate data	Assessment-only, reproducible	Re-run assessment
NSG flow logs	Diagnostic data	Not critical for recovery

DR Procedure

1. Assess the failure scope — single resource vs. full region outage
2. For single resource failure — use the Operations Runbook procedures
3. For full region failure:
   • Target the failover region: azd env set AZURE_LOCATION germanywestcentral
   • Update CIDR ranges if needed
   • Run azd up to redeploy all infrastructure
   • Restore VMs from the Recovery Services Vault (if LRS, data is region-local; for cross-region DR, upgrade to GRS)

Caution

With LRS storage redundancy, backup data does not replicate to the secondary region. For cross-region DR, upgrade the Recovery Services Vault to GRS (~2x backup storage cost).