VMware Cloud Foundation Deployment with Integrated DR: A Design-Centric Implementation Approach

Introduction:

Architectural Intent First, Tools Second

VMware Cloud Foundation (VCF) is not a product — it is a converged infrastructure framework. It standardizes the full software-defined data center stack through automated lifecycle management, policy-driven operations, and intrinsic security.

Designing and deploying VCF from scratch is a high-stakes operation, particularly when disaster recovery (DR) is not an afterthought, but a built-in objective. This article captures the real-world implementation strategy for VCF 4.x+ including its integration with Site Recovery Manager (SRM) and vSphere Replication, covering both stretched and active/passive designs.

1. VCF Architecture – Core Principles

a. VCF Domains:

• Management Domain: Dedicated for infrastructure services (vCenter, NSX-T, SDDC Manager, vSAN, etc.)
• Workload Domains (WLD): Application/service boundaries — each domain gets its own vCenter, NSX, and vSAN if needed.

b. Integrated Stack:

• vSphere for compute
• vSAN as the native storage layer
• NSX-T for network virtualization and security
• SDDC Manager for lifecycle automation, updates, and configuration drift enforcement

c. Deployment Toolchain:

• Cloud Builder Appliance initiates the bring-up of the management domain using imaging bundles and config JSON.
• LCM via SDDC Manager handles version control, upgrade planning, and patch enforcement across all components.

2. Pre-Deployment Design Considerations

A. BOM + HCL Alignment

• All hardware (compute, NICs, SSDs, RAID controllers) must be listed in the VCF-certified BOM and VMware HCL.

b. Networking

• L2/L3 VLANs: Isolate Management, vMotion, vSAN, NSX overlay, and Edge uplinks
• MTU 9000 is mandatory end-to-end
• Redundant physical switches with LAG or vPC

c. NTP, DNS, Certificates

• Internal or external DNS must resolve FQDNs pre-deployment
• All services must be time-synced (NTP) or drift will break SDDC Manager functionality
• Use custom signed certs or deploy with VMCA and replace post-implementation

3. Implementation Workflow (Phased)

a. Phase 1: Imaging

• Prepare ESXi hosts with matching firmware
• Use Imaging bundle from VMware or vLCM JSON profile

b. Phase 2: Management Domain Bring-Up

• Deploy Cloud Builder
• Validate JSON config (cluster name, IPs, FQDNs, domain, NTP, DNS, passwords)
• Deploy management VMs (vCenter, NSX-T Manager, SDDC Manager, vSAN Cluster)

c. Phase 3: Workload Domain Creation

• Use SDDC Manager to instantiate additional WLDs
• Choose vSAN vs NFS
• Automate network provisioning via NSX-T

4. DR Design Integration

a. Option 1: SRM with vSphere Replication (Active-Passive)

• Two independent VCF instances (Primary & DR)
• SRM deployed on each management domain
• Replication via vSphere Replication or array-based (PowerMax, vVOLs, etc.)
• DNS failover required or use GSLB
• NSX-T must be manually extended or stretched

b. Option 2: Stretched Cluster (Active-Active)

• vSAN Stretched across two fault domains
• Witness appliance placed in 3rd location
• NSX-T segments stretched manually (beware of BUM traffic optimization)
• Metro clustering only valid with <5ms RTT
• Quorum must be preserved (split-brain risk mitigated)

5. Operationalization & Lifecycle

• Use SDDC Manager LCM bundles to enforce consistent patch levels
• Backup SDDC Manager, vCenter, NSX-T, and vSAN separately
• Use Aria Suite (vROps, vRLI, vRA) for monitoring, logging, and automated remediation
• DR testing via SRM Recovery Plans — validate with runbook automation

6. Real-World Issues to Avoid

Issue	Impact	Resolution
Incomplete DNS resolution	Cloud Builder fails	Pre-stage all records incl. reverse lookup
Mixed firmware levels	Bring-up failure	Use HCL-validated vLCM image
Misaligned NSX VLANs	Overlay drops or fails to route	Pre-validate physical underlay mapping
Storage Policy mismatch	vSAN object inaccessible	Sync storage policies across WLDs

📌 Final Thoughts

Deploying VMware Cloud Foundation from scratch is not just a task — it’s a commitment to architectural standardization, lifecycle automation, and design discipline. When disaster recovery is embedded into the blueprint, not added later, the result is a truly resilient SDDC.

VCF + DR Architecture Diagram

The following diagram illustrates the high-level design for a dual-region VCF deployment with integrated SRM-based disaster recovery

✍️ Author

Mohamed Omar — Infrastructure Architect and VMware Consultant with 17+ years in SDDC, vSAN, VCF, DR, and enterprise-grade virtualization architecture.