Server relocation to a data center: plan, tests and rollback

An operational server relocation plan: inventory, backup, service window, DNS, IP addressing, transport, acceptance tests, go/no-go criteria and rollback.

Problem: infrastructure move

Server relocation to a data center: plan, tests and rollback

Server relocation is not just transport. The project should have named decision owners, a complete inventory, verified backup, acceptance tests and explicit criteria for continuation or rollback.

Inventory before the change

A list of servers, services, ports, VLANs, public IP addresses, DNS, certificates, backups and application dependencies helps define the relocation sequence.

Service window and rollback

Before transport, lower DNS TTL, prepare snapshots or backups, define rollback criteria and assign people responsible for the go/no-go decision.

Launch in the new data center

After installation, verify power, uplinks, routing, firewall, VPN, monitoring, administrative access, mail, websites and business applications.

Hybrid model during migration

Some services can temporarily run on VPS, Cloud Pro or a dedicated server to reduce risk and shorten application downtime.

Six-stage relocation plan

Each stage has a tangible output and a decision gate. The exact service window and scope are confirmed only after the inventory.

01

Discovery and inventory

Servers, power draw, rack units, ports, VLANs, IP addresses, DNS, certificates, dependencies, owners and startup order.

02

Target design

Rack layout, A/B power, uplinks, addressing, routing, firewall, VPN, monitoring, backup and access paths.

03

Rehearsal and readiness

Backup restore check, lower DNS TTL, prepared configurations, contact tree, transport plan and agreed go/no-go criteria.

04

Service window and transport

Controlled shutdown, equipment identification, protected transport, installation, cabling and power-on in the agreed sequence.

05

Cutover and acceptance

Network, DNS, firewall, VPN, application, database, mail, monitoring, backup and end-user validation.

06

Stabilisation and closure

Enhanced monitoring, incident log, documentation update, final acceptance and closure of the rollback window.

Acceptance tests and rollback gates

Area
Acceptance evidence
Stop or rollback condition
Infrastructure
A/B power, cabling, uplinks and remote access are operational.
Stop if redundant power or the required management path is unavailable.
Network
Expected routes, VLANs, DNS, firewall, VPN and public services pass tests.
Rollback if critical routes or security policy cannot be restored in the window.
Applications
Health checks, databases, mail, APIs and user scenarios complete successfully.
Rollback when a critical business process fails and no accepted workaround exists.
Data and backup
Data consistency is confirmed and a restore path is available.
Stop when backup integrity or required recovery evidence is missing.
Monitoring
Metrics, logs and alerts arrive from all critical systems.
Do not close the window while critical systems remain unobservable.

Responsibility matrix

Workstream
Owner
Decision or deliverable
Application and business owners
Client
Priorities, dependencies, user acceptance and business go/no-go.
Data center, power and connectivity
DataHouse / eTop
Prepared rack space, agreed power, connectivity and physical access.
Configuration and cutover
Agreed project team
Exact responsibility for DNS, firewall, systems and applications is set in the migration plan.
Final acceptance and rollback
Joint decision
Named technical and business owners decide from documented test results.
What we need to prepare an individual plan: equipment list and power draw, rack space, current and target addressing, connectivity, service dependencies, required service window, backup status and technical contacts. DataHouse confirms the final scope, tests and responsibilities in the project documentation.

Plan the relocation with DataHouse. Send the inventory and expected date. We will identify dependencies, define the migration sequence and prepare the acceptance and rollback framework.

Request a migration plan

Technical owner: DataHouse / eTop team. Reviewed: 27 July 2026. Exact commercial and SLA parameters are confirmed in the offer and agreement.

Decision signals

Best fit

companies moving own servers, racks, ERP, mail, databases, storage and workloads from an office server room

Key risks

downtime, DNS, IP addressing, firewall rules, backup, transport, application dependencies and post-launch tests

DataHouse services

colocation, dedicated servers, Cloud Pro, VPS, administration, monitoring, backup and diagnostics

Search intent

server relocation, move servers to data center, data center migration, colocation migration

Frequently asked questions

Does server relocation require IP address changes?

Not always. It depends on the current addressing, provider and service model. The project should plan routing, DNS, firewall and post-cutover tests.

How can downtime be reduced during relocation?

Backups, lower DNS TTL, prepared firewall rules, ready rack space, power tests and a clear rollback plan help reduce downtime.

Can relocation be combined with cloud migration?

Yes. Some services can move to colocation and others to VPS, Cloud Pro or dedicated servers when this reduces operational risk.

What should the pre-relocation inventory include?

Servers and power draw, ports, VLANs, IP addresses, DNS, certificates, application dependencies, contacts, backups and the required startup sequence.

Who makes the go/no-go decision?

A business owner and a technical owner should be named before the service window. The decision should use evidence about backup, connectivity, rack readiness, teams and rollback.

Which tests are run after launch?

At minimum: power, uplinks, routing, DNS, firewall, VPN, monitoring, administrative access, applications, databases, mail, backup and an end-user test.