Data Center Relocation: The Complete Guide

Data Center Relocation: Why It Requires a Different Approach

IT infrastructure relocation at the data center level is a categorically different undertaking from a standard office IT move. An office IT move is an infrastructure that helps a set of users. A data center relocation project can involve infrastructure that serves thousands of users, dozens of external clients, and compliance requirements in multiple jurisdictions, all while preserving availability requirements that underpin your business.

There are different failure modes, too. Downtime in an office move is instant and on-site. In the event of a data center relocation, failures ripple through interconnected systems, which can take hours to identify, and can involve external customers that may sense it first, well before your team. That is why many businesses prefer to outsource data center relocation to a specialist data center relocation companies instead of handling it themselves.

Phase 1: Infrastructure Discovery and Documentation

Complete and accurate documentation is the foundation of every other phase. This usually takes 4-8 weeks in a complex environment and cannot be shortened without putting at risk. Use it as the starting point for your data center moving checklist and call a qualified data center moving company early to confirm your inventory before any hardware moves.

• Complete physical asset, inventory every server, storage device, network appliance, and cable
• Logical architecture documentation, network topology, VLAN design, routing, firewall zones
• Application dependency mapping, which applications depend on which infrastructure
• Performance baseline, current utilization, and performance metrics for all critical systems
• Connectivity inventory, all external circuits, peering relationships, CDN configurations, API integrations
• Compliance inventory, systems subject to PCI-DSS, HIPAA, PIPEDA, SOX

Phase 2: Migration Strategy Selection

The most critical decision in the whole server migration project is the choice of the server migration strategy. The three approaches listed below are recommended by data center relocation companies with experience, based on the level of risk, budget, and complexity of the environment.

Option A: Lift and Shift

Physical servers shut down, are transported, and are restarted at the destination. Highest downtime risk. Useful only if the destination infrastructure is the same as the source one, and the organization has a desired maintenance period. Low cost, high risk.

Option B: Phased Migration with Parallel Operations

Migrations were accomplished in stages over the course of weeks, with both locations running concurrently. All workloads are migrated, tested, and validated before the source is decommissioned. Increased expenses due to operating at two locations. Recommended for complex environments or business-critical environments. A data center moving company that you can trust can manage to coordinate the moving of all the physical information between the two data centers, allowing your team to focus on the process of validating and ensuring continuity.

Option C: Cloud Migration Opportunity

A data center relocation is one of the best opportunities to evaluate whether workloads should move to the cloud rather than new physical infrastructure. The planning work required for physical migration is largely reusable for cloud assessment. Evaluate this option early, as it affects the entire migration architecture.

Phase 3: Destination Preparation

No data center relocation should proceed to live migration until the destination facility is fully prepared. A reputable data center moving companies considers destination readiness to be one of the items on the data center moving checklist, and nothing moves until each of the following items is checked off:

• Power and cooling confirmed and tested under load simulation
• Network infrastructure deployed and configured to match the source VLAN design
• External connectivity established, internet circuits, private WAN, and carrier connections
• Security controls deployed, firewalls, IDS/IPS, and access control
• Monitoring operational systems before production systems arrive

Phase 4: Tiered Migration Execution

• Development and test environments: Low risk and high learning value. Use these to validate procedures before touching production.
• Non-critical production systems: Back-office, analytics, archival. If not immediately business critical, more consequences than test.
• Supporting production infrastructure: Authentication, DNS, monitoring, backup. Must be running at the destination before migrating the production applications.
• Production applications: Last to be migrated, with validated processes and confirmed destination infrastructure.

Phase 5: Cutover and Post-Migration Validation

The final cutover, migrating the last production workloads and decommissioning the source, is the highest-risk event in any data center relocation. It is a formal event with documented go/no-go criteria and a 72-hour hypercare window. During this period, data center relocation companies have teams on-site to handle any problems as they occur:

• Cutover window during the minimum-traffic period, typically a weekend night
• Rollback procedures documented and tested before cutover begins
• External DNS TTLs were reduced 48 hours before cutover to enable fast failback
• Performance compared against the pre-migration baseline for every critical system
• The source environment is kept on standby for 72 hours before decommission begins