Architecture patterns

These patterns cover the most common production deployments of Datamotive. Each pattern describes the topology, the products involved, and the network and storage requirements.

Pattern 1: VMware to cloud DR

A single on-premises VMware cluster replicates to a public cloud target (AWS or Azure). This is the most common starting point for organizations moving their DR strategy off a secondary physical data center.

Components:

• Source: VMware vSphere 6.7 or later
• Target: AWS EC2 or Azure VMs
• Control plane: Datamotive SaaS-managed or self-hosted
• Datamotive Replication Appliance: one VM per vSphere cluster

How replication works:

1. The Replication Appliance reads changed blocks from the vSphere snapshot API (no agent on VMs)
2. Changed blocks are compressed, encrypted, and transferred to the target cloud region
3. A shadow VM in the target cloud is kept in a powered-off, ready-to-boot state
4. On failover, the shadow VM boots and network mappings are applied

Network requirements:

• TCP 443 outbound from the Replication Appliance to the target cloud and Datamotive control plane
• Bandwidth: 10 Mbps sustained per 1 TB of daily changed data (approximate)

Pattern 2: Cross-cloud DR (cloud to cloud)

A workload running in one public cloud region replicates to a different cloud provider or region. This pattern removes vendor lock-in from the DR strategy.

Example topologies:

• AWS us-east-1 to Azure East US
• Azure West Europe to AWS eu-west-1
• AWS primary to AWS DR in a different region (for compliance isolation)

Components:

• Source: EC2 or Azure VMs
• Target: Any supported cloud platform
• Datamotive Cloud Connector deployed in each cloud account

Key consideration: Data transfer costs apply between cloud providers. Datamotive compresses blocks before transfer, which typically reduces egress by 30 to 60 percent depending on data type.

Pattern 3: Multi-site DR (hub-and-spoke)

Multiple source sites (branches, data centers, cloud regions) replicate to a single shared DR target. A central Datamotive control plane manages all replication plans.

Topology:

• N source sites, each with a Replication Appliance or Cloud Connector
• 1 target DR site (private cloud or public cloud)
• 1 Datamotive control plane instance (SaaS or self-hosted)

Use case: Retail chains, financial services branches, or multi-campus universities where each site has critical workloads but a shared DR budget.

Pattern 4: Migration wave

A phased migration project where workloads are synchronized to the target and cut over in waves, with each wave running over a weekend or maintenance window.

Wave structure:

1. Discovery and grouping: classify VMs by application dependency
2. Pre-sync: start continuous replication weeks before the cutover
3. Cutover window: at T-0, quiesce the source, trigger final sync, bring up the target
4. Validation: smoke-test applications on the target before decommissioning the source

Datamotive product: Easy Migrate handles wave orchestration. Each wave is a migration plan that groups dependent VMs and tracks cutover status.

Pattern 5: Layered protection (DR plus ransomware)

Easy Hybrid DR provides the baseline RTO. Easy Protect adds ransomware detection, scanning each replication cycle for anomalies and maintaining a verified clean-state recovery point.

When to use: Tier 1 systems in healthcare, BFSI, or manufacturing where both sub-10-minute RTO and clean-state integrity are required simultaneously.

Related docs

• Prerequisites

• Sizing guidelines

• Logical architecture