Skip to content

Azure Local Disconnected Mode



Disconnected Mode enables Azure Local clusters to operate completely independently without continuous connectivity to Microsoft Azure or the internet. This mode is designed for air-gapped environments, classified networks, and scenarios where data must remain completely isolated.

graph TB
    subgraph AirGap[🔒 Air-Gapped Environment - No External Connectivity]
        subgraph Local[On-Premises Infrastructure]
            Cluster[Azure Local Cluster<br/>Fully Autonomous]
            WAC[Windows Admin Center<br/>Local Management]
            VMs[Virtual Machines]
            Storage[Storage Spaces Direct]
            Backup[Local Backup]
        end

        Admin[Administrator] --> WAC
        WAC --> Cluster
        Cluster --> VMs
        Cluster --> Storage
        VMs --> Backup

        style Local fill:#FFF4E6,stroke:#FF8C00,stroke-width:3px,color:#000
    end

    Azure[Azure Cloud<br/>❌ No Connection]
    Internet[Internet<br/>❌ No Access]

    style AirGap fill:#F8F8F8,stroke:#D13438,stroke-width:3px,stroke-dasharray: 5 5,color:#000
    style Azure fill:#E8F4FD,stroke:#999,stroke-width:1px,stroke-dasharray: 3 3,color:#999
    style Internet fill:#F3E8FF,stroke:#999,stroke-width:1px,stroke-dasharray: 3 3,color:#999
    style Storage fill:#D4E9D7,stroke:#107C10,stroke-width:2px,color:#000

Connectivity:

  • No continuous Azure connectivity required
  • Optional periodic connectivity for updates
  • Can be completely air-gapped indefinitely
  • All management functions available locally

Management:

  • Windows Admin Center for local management
  • PowerShell for automation and scripting
  • No Azure portal access
  • All operations are on-premises

Operations:

  • Fully autonomous cluster operation
  • No dependency on cloud services
  • Local backup and DR solutions
  • Manual update deployment
  • Complete operational sovereignty

← Back to Azure Local Overview


Definition: An air-gapped environment has no physical or logical connection to external networks, including the internet.

Common Scenarios:

  • Classified Government Networks: Defense, intelligence, national security
  • Secure Research Facilities: Sensitive R&D, weapons research, national labs
  • Critical Infrastructure: Power plants, water treatment, nuclear facilities
  • Financial Trading Floors: High-frequency trading systems
  • Healthcare Isolated Networks: Psychiatric facilities, research hospitals

Compliance Drivers:

  • ITAR (International Traffic in Arms Regulations)
  • National security requirements
  • Top Secret/SCI classifications
  • SCIF (Sensitive Compartmented Information Facility)
  • Critical infrastructure protection regulations

Disconnected Mode is also appropriate for:

Highly Regulated Industries:

  • Banking core systems (SOX, PCI-DSS)
  • Healthcare clinical systems (HIPAA)
  • Government citizen data (privacy laws)
  • Legal case management (attorney-client privilege)

Operational Requirements:

  • Zero tolerance for internet dependency
  • Complete control over all data flows
  • No external communication allowed
  • Isolation from corporate networks

Threat Models:

  • Protection against advanced persistent threats (APT)
  • Prevention of data exfiltration
  • Protection against supply chain attacks
  • Insider threat mitigation

Remote Locations:

  • Oil rigs and offshore platforms
  • Remote military bases
  • Polar research stations
  • Ships and submarines
  • Mining operations

Unreliable Connectivity:

  • Locations with frequent outages
  • High-latency satellite connections
  • Bandwidth-constrained environments
  • Networks with unpredictable availability

Some organizations choose Disconnected Mode to:

  • Avoid cloud service costs
  • Reduce WAN bandwidth requirements
  • Simplify licensing compliance
  • Minimize external dependencies

Azure Integration:

  • ❌ Azure portal management
  • ❌ Azure Arc registration
  • ❌ Azure Monitor integration
  • ❌ Azure Backup to cloud
  • ❌ Azure Site Recovery
  • ❌ Azure Security Center
  • ❌ Azure Policy enforcement
  • ❌ Azure Update Manager

Hybrid Services:

  • ❌ Azure Virtual Desktop (cloud-managed)
  • ❌ Azure IoT Hub integration
  • ❌ Azure Machine Learning cloud training
  • ❌ Azure AD authentication (cloud-only)
  • ❌ Automatic updates from Microsoft

Observability:

  • ❌ Centralized monitoring across clusters
  • ❌ Cloud-based log analytics
  • ❌ Cross-cluster dashboards
  • ❌ Azure-based alerting

Core Capabilities:

  • ✅ All VM workloads
  • ✅ Storage Spaces Direct
  • ✅ Software-defined networking
  • ✅ Hyper-V and AKS
  • ✅ High availability and failover
  • ✅ Local backup solutions
  • ✅ GPU acceleration
  • ✅ Edge AI/ML workloads

Management:

  • ✅ Windows Admin Center (local)
  • ✅ PowerShell automation
  • ✅ Local Active Directory
  • ✅ Group Policy
  • ✅ Local RBAC
  • ✅ Event log monitoring
  • ✅ Performance monitoring

Security:

  • ✅ BitLocker encryption
  • ✅ Secure Boot and TPM
  • ✅ Network isolation
  • ✅ Local firewall
  • ✅ Windows Defender
  • ✅ Local security policies

Comparison Table:

FeatureConnected ModeDisconnected Mode
VM Workloads✅ Full✅ Full
Storage Performance✅ Full✅ Full
Networking✅ Full✅ Full
High Availability✅ Yes✅ Yes
Azure Portal✅ Yes❌ No
Azure Monitor✅ Yes❌ No (local only)
Cloud Backup✅ Yes❌ No
Auto Updates✅ Yes❌ Manual only
Cost🟡 Higher🟢 Lower
Complexity🟢 Lower🟡 Higher
Air-gap Support❌ No✅ Yes

Even in Disconnected Mode, periodic synchronization can be beneficial when security policies permit.

Option 1: Periodic Network Connection

  • Connect to internet/Azure on schedule (for example, monthly)
  • Download updates and telemetry
  • Sync configuration changes
  • Disconnect after sync complete

Frequency:

  • Monthly (most common)
  • Quarterly (for very restrictive environments)
  • On-demand (when specific updates needed)

Option 2: Air-Gapped Transfer

  • Download updates to external media
  • Physical transfer to disconnected environment
  • Apply updates locally
  • No network connection required

Process:

  1. Use connected system to download updates
  2. Copy to USB drive or portable hard drive
  3. Security scan media before import
  4. Transfer to disconnected environment
  5. Import and apply updates
  6. Document transfer in audit log

Option 3: One-Way Data Diode

  • Hardware-enforced unidirectional data flow
  • Allows updates in, prevents data out
  • Used in high-security environments
  • Expensive but provides maximum assurance

Updates:

  • Windows Server cumulative updates
  • Azure Local platform updates
  • Security patches
  • Driver and firmware updates

Telemetry (Optional):

  • Health and diagnostic data
  • Performance metrics
  • Fault reports
  • For vendor support purposes

Configuration:

  • Policy updates
  • Certificate renewals
  • License activations

Between Sync Periods:

Baseline Configuration:

  • Establish approved configuration baseline
  • Document any deviations
  • Review during next sync
  • Restore compliance if needed

Change Control:

  • Strict change management process
  • Document all changes
  • Review and approve before implementation
  • Audit regularly

Drift Detection:

  • Regular configuration audits
  • Compare against baseline
  • Identify unauthorized changes
  • Remediate before next sync

Management Without Continuous Cloud Connection

Section titled “Management Without Continuous Cloud Connection”

Primary Management Interface:

Capabilities:

  • Full cluster management
  • VM lifecycle operations
  • Storage configuration
  • Network management
  • Performance monitoring
  • Update deployment
  • Certificate management

Access Methods:

  • Installed on management workstation
  • Installed on cluster nodes (gateway mode)
  • Remote access via RDP/VPN
  • High availability deployment

Advantages:

  • Rich graphical interface
  • No cloud dependency
  • Real-time operations
  • Comprehensive feature set

Reference: Windows Admin Center

Comprehensive Automation:

Common Tasks:

Terminal window
# Cluster management
Get-ClusterNode
Get-ClusterResource
Get-ClusterSharedVolume
# VM operations
Get-VM
New-VM -Name "AppServer01" -MemoryStartupBytes 8GB -VHDPath "C:\VMs\App01.vhdx"
Start-VM -Name "AppServer01"
# Storage management
Get-StoragePool
Get-Volume
New-Volume -FriendlyName "Data01" -Size 1TB -ResiliencySettingName "Mirror"
# Network management
Get-VMSwitch
New-VMSwitch -Name "ExternalSwitch" -NetAdapterName "Ethernet"
# Health monitoring
Get-HealthFault
Get-StorageJob

Automation Scripts:

  • Scheduled health checks
  • Backup automation
  • Report generation
  • Configuration validation

Event-Based Monitoring:

Windows Event Logs:

  • System logs for cluster events
  • Application logs for workloads
  • Security logs for auditing
  • Forwarding to SIEM (if available)

Performance Monitor:

  • Real-time performance metrics
  • Historical data collection
  • Custom counters and alerts
  • Performance baselines

Health Service:

  • Built-in Azure Local health monitoring
  • Predictive failure detection
  • Storage capacity alerts
  • Performance degradation detection

Custom Monitoring Solutions:

  • SCOM (System Center Operations Manager)
  • Nagios, Zabbix, Prometheus
  • Custom PowerShell scripts
  • Email/SMS alerting

Manual Update Process:

Monthly Update Cycle:

  1. Preparation (Week 1):

    • Review Microsoft security bulletin
    • Download updates to test environment
    • Test in non-production cluster
    • Document results
  2. Change Control (Week 2):

    • Submit change request
    • Get approval from stakeholders
    • Schedule maintenance window
    • Communicate to users
  3. Deployment (Week 3):

    • Backup cluster configuration
    • Run Cluster-Aware Updating (CAU)
    • Monitor deployment progress
    • Validate after each node
  4. Validation (Week 4):

    • Full cluster health check
    • Application testing
    • Performance validation
    • Document completion

Emergency Patching:

  • Expedited process for critical security patches
  • Out-of-band deployment
  • Shortened testing cycle
  • Enhanced monitoring post-deployment

Disconnected Mode provides inherent security advantages:

Attack Surface Reduction:

  • No internet-facing services
  • No inbound connections from external networks
  • Eliminates entire classes of attacks
  • Reduced vulnerability to zero-day exploits

Data Exfiltration Prevention:

  • Physical air-gap prevents data leakage
  • No path for malware to “phone home”
  • Insider threats limited to physical access
  • Command and control (C2) channels blocked

Supply Chain Protection:

  • Controlled update process
  • Verification of all imported code
  • Malicious update prevention
  • Reduced risk of compromised dependencies

Baseline Security:

Operating System:

  • CIS Benchmark Level 2
  • STIG (Security Technical Implementation Guide)
  • Minimal installation
  • Disabled unnecessary services
  • Regular vulnerability scanning

Network Security:

  • Host-based firewall enabled
  • Network segmentation (VLANs)
  • Strict access control lists
  • Intrusion detection systems (IDS)
  • No wireless networks

Access Controls:

  • Multi-factor authentication (MFA)
  • Privileged access management (PAM)
  • Just-in-time admin access
  • Regular access reviews
  • Separation of duties

Critical for disconnected environments:

Hardware Security:

  • Locked racks and enclosures
  • Tamper-evident seals
  • Serial number tracking
  • Secure disposal procedures
  • No unauthorized devices

Facility Security:

  • Badge access control
  • Video surveillance
  • Security guards
  • Visitor logs and escorts
  • Mantraps and dual authentication

Media Security:

  • Secure storage for update media
  • Chain of custody for transfers
  • Media sanitization procedures
  • Encrypted storage
  • Regular media inventory

Audit Logging:

  • Comprehensive event logging
  • Log forwarding to SIEM
  • Tamper-proof log storage
  • Regular log review
  • Long-term retention (7+ years)

Compliance Frameworks:

  • NIST 800-53 (government)
  • ITAR (defense industry)
  • HIPAA (healthcare)
  • PCI-DSS (payment cards)
  • ISO 27001 (information security)

Regular Audits:

  • Quarterly security assessments
  • Annual penetration testing
  • Configuration audits
  • Access reviews
  • Vulnerability scanning

Documentation:

  • System security plan (SSP)
  • Security controls documentation
  • Risk assessment
  • Incident response plan
  • Business continuity plan

Problem:

  • Time-consuming process
  • Risk of human error
  • Delayed patch deployment
  • Testing overhead

Mitigation Strategies:

  • Establish documented procedures
  • Automate with PowerShell where possible
  • Maintain test environment
  • Use CAU for orchestration
  • Schedule regular update cycles
  • Prioritize critical security patches

Problem:

  • No centralized monitoring across clusters
  • Manual log collection
  • Difficult to correlate events
  • Reactive rather than proactive

Mitigation Strategies:

  • Deploy local SIEM solution
  • Implement automated health checks
  • Use PowerShell for regular reporting
  • Establish performance baselines
  • Create custom dashboards in Windows Admin Center
  • Document normal operational patterns

Problem:

  • No cloud-based backup
  • Local infrastructure required
  • Offsite protection challenges
  • Recovery time objectives

Mitigation Strategies:

  • Deploy secondary Azure Local cluster for local DR
  • Use traditional backup solutions (Veeam, Commvault)
  • Replicate to secure offsite facility
  • Regular restore testing
  • Automate backup monitoring
  • Implement backup verification

Challenge 4: Troubleshooting Without Cloud Support

Section titled “Challenge 4: Troubleshooting Without Cloud Support”

Problem:

  • No Azure support diagnostics
  • Limited remote support options
  • Extended resolution times

Mitigation Strategies:

  • Build strong internal expertise
  • Maintain comprehensive documentation
  • Establish peer support networks
  • Use Microsoft Premier Support effectively
  • Collect diagnostic data for offline analysis
  • Create detailed runbooks

Problem:

  • Certificates expire without internet access
  • Renewal process is manual
  • Risk of expired certificates causing outages

Mitigation Strategies:

  • Implement internal Certificate Authority (CA)
  • Automate certificate monitoring
  • Alert 90 days before expiration
  • Document renewal procedures
  • Test certificate renewal process
  • Use longer validity periods (where allowed)

Problem:

  • Activation and licensing verification
  • Volume licensing management
  • Offline activation requirements

Mitigation Strategies:

  • Use Volume Licensing (VL) with KMS
  • Implement local Key Management Service
  • Document activation procedures
  • Plan for periodic re-activation
  • Maintain license documentation
  • Work with Microsoft licensing team

Use Case 1: Defense Agency Classified Network

Section titled “Use Case 1: Defense Agency Classified Network”

Scenario:

  • Top Secret/SCI classified network
  • Complete air-gap required (no internet)
  • 500+ VMs across 8-node cluster
  • Mission-critical intelligence applications
  • Zero tolerance for data spillage

Disconnected Mode Implementation:

  • No Azure connectivity ever
  • Windows Admin Center for management
  • Manual quarterly updates via secure media
  • Local backup to secondary cluster
  • Physical security controls
  • Comprehensive audit logging

Results:

  • ✅ 100% compliance with security requirements
  • ✅ Complete operational sovereignty
  • ✅ No risk of data exfiltration
  • ✅ 99.9% uptime for 3 years
  • ✅ Passed all security audits
  • ✅ Zero security incidents

Key Learnings:

  • Quarterly updates sufficient for security
  • Local expertise critical for operations
  • Documentation and runbooks essential
  • Regular testing of DR procedures mandatory

Use Case 2: Nuclear Power Plant Control Systems

Section titled “Use Case 2: Nuclear Power Plant Control Systems”

Scenario:

  • Nuclear power generation facility
  • Critical infrastructure protection requirements
  • Real-time SCADA and control systems
  • Air-gapped from corporate and internet
  • NRC (Nuclear Regulatory Commission) compliance

Disconnected Mode Implementation:

  • Completely isolated network
  • 4-node cluster for redundancy
  • Local management only
  • Monthly updates via secure process
  • Dedicated operations team
  • Offline backup and DR

Results:

  • ✅ NRC compliance maintained
  • ✅ Zero unplanned downtime
  • ✅ Protection against cyber threats
  • ✅ Rapid response to operational needs
  • ✅ Complete audit trail
  • ✅ Successful disaster recovery tests

Key Learnings:

  • Physical security as important as cyber
  • Change control critical for safety
  • Regular training for operations staff
  • Simulate failure scenarios regularly

Use Case 3: Financial High-Frequency Trading

Section titled “Use Case 3: Financial High-Frequency Trading”

Scenario:

  • Trading floor with HFT algorithms
  • Sub-millisecond latency requirement
  • Isolated from corporate network
  • Proprietary algorithms (IP protection)
  • SEC regulatory requirements

Disconnected Mode Implementation:

  • Isolated VLAN segment
  • No internet or corporate connectivity during market hours
  • 2-node cluster with GPU acceleration
  • After-hours sync for updates
  • Real-time risk monitoring (local)
  • Comprehensive audit logging

Results:

  • ✅ Sub-500 microsecond latency
  • ✅ Zero IP leakage incidents
  • ✅ SEC compliance maintained
  • ✅ 99.99% uptime during market hours
  • ✅ Protected proprietary algorithms
  • ✅ Successful regulatory audits

Key Learnings:

  • Network isolation critical for latency
  • After-hours maintenance window sufficient
  • Local monitoring essential for trading
  • Disaster recovery must be seconds, not minutes

Use Case 4: Pharmaceutical Research Laboratory

Section titled “Use Case 4: Pharmaceutical Research Laboratory”

Scenario:

  • Drug discovery research
  • Sensitive compound data (IP protection)
  • FDA compliance requirements
  • Collaboration with external partners (limited)
  • High-performance computing for simulations

Disconnected Mode Implementation:

  • Air-gapped research network
  • 8-node cluster with GPU nodes
  • Data import via secure transfer process
  • Local AI/ML model training
  • Monthly external sync for approved data
  • Comprehensive data classification

Results:

  • ✅ Complete IP protection (zero breaches)
  • ✅ FDA 21 CFR Part 11 compliance
  • ✅ 50% faster drug discovery cycle
  • ✅ Successful collaboration with partners
  • ✅ Data sovereignty maintained
  • ✅ Audit-ready documentation

Key Learnings:

  • Data classification essential
  • Secure data transfer process critical
  • GPU acceleration enables faster research
  • Periodic sync sufficient for collaboration

Scenario:

  • Remote mining site in Australia outback
  • No reliable internet connectivity
  • Operations cannot stop for connectivity
  • Safety-critical monitoring systems
  • 200+ VMs for various systems

Disconnected Mode Implementation:

  • Complete offline operation
  • Satellite link for emergency only
  • 4-node cluster
  • Local management team
  • Quarterly updates via physical media
  • Diesel generators for power

Results:

  • ✅ Zero downtime due to connectivity
  • ✅ Continuous mining operations
  • ✅ Safety systems always available
  • ✅ Cost savings (no WAN costs)
  • ✅ Autonomous operations
  • ✅ Improved operational efficiency

Key Learnings:

  • Disconnected mode enables remote operations
  • Local expertise required for maintenance
  • Power and cooling more challenging than connectivity
  • Regular maintenance schedule critical

1. Comprehensive Baseline:

  • Document approved configuration
  • Create detailed architecture diagrams
  • Establish security baselines
  • Define operational procedures
  • Document all dependencies

2. Test Environment:

  • Mirror production configuration
  • Test all updates before deployment
  • Practice DR scenarios
  • Validate backup/restore
  • Document test results

3. Change Management:

  • Formal change control process
  • Risk assessment for all changes
  • Approval workflow
  • Rollback procedures
  • Post-implementation review

1. Regular Health Checks:

  • Daily automated health reports
  • Weekly performance review
  • Monthly capacity planning
  • Quarterly security scans
  • Annual full audit

2. Documentation:

  • Keep all documentation current
  • Maintain runbooks for common tasks
  • Document all configuration changes
  • Create troubleshooting guides
  • Regular documentation review

3. Training:

  • Train multiple staff members
  • Cross-train for redundancy
  • Regular knowledge refresh
  • Simulate failure scenarios
  • Document lessons learned

4. Backup and DR:

  • Daily backups (automated)
  • Weekly full backups
  • Offsite backup copies
  • Monthly DR tests
  • Documented recovery procedures

1. Defense in Depth:

  • Multiple security layers
  • Network segmentation
  • Least privilege access
  • Regular security assessments
  • Continuous monitoring

2. Access Control:

  • Strong authentication (MFA)
  • Regular access reviews
  • Privileged access management
  • Session recording
  • Audit all access

3. Physical Security:

  • Locked data center
  • Video surveillance
  • Badge access
  • Visitor escorts
  • Equipment inventory

4. Incident Response:

  • Documented IR plan
  • Regular IR drills
  • Contact information current
  • Escalation procedures
  • Post-incident review

Comparison: When to Choose Disconnected Mode

Section titled “Comparison: When to Choose Disconnected Mode”

✅ Strong Yes:

  • Air-gap requirement (regulatory or policy)
  • Classified or top-secret data
  • Critical infrastructure protection
  • Zero tolerance for internet dependency
  • Complete operational sovereignty required
  • Protection against advanced persistent threats

⚠️ Consider Carefully:

  • Staff expertise to manage without cloud support
  • Budget for local infrastructure (backup, monitoring)
  • Change management discipline
  • Physical security capabilities
  • Compliance requirements justify the complexity

✅ Better Choice:

  • No air-gap requirement
  • Centralized management desired
  • Cloud backup and DR beneficial
  • Limited local IT staff
  • Multiple clusters to manage
  • Cost optimization via cloud services

✅ Best of Both:

  • Start in Connected Mode for deployment and validation
  • Switch to Disconnected Mode for production
  • Periodic sync (monthly/quarterly)
  • Emergency reconnect capability
  • Balance sovereignty with manageability

Continue Learning:

Back to:

External Resources:


Last Updated: October 2025