IT relocation checklist

This is the physical move only. It doesn't cover application migration, data replication or network design. Your technical team looks after those, alongside this list.

8 to 12 weeks before: plan and scope

• Name a business owner and a technical lead.
• Agree why you're moving, what's in scope and the target date.
• Choose how you'll move: all at once, in phases, or with spare kit to cover the gap.
• Agree the downtime window with the business.
• Survey both sites, and the route from the rack to the vehicle at each end.
• Appoint your mover.
• Check the cover for the equipment in transit.
• Start a list of risks and dependencies.

4 to 8 weeks before: audit and design

• List every asset with make, model and serial number.
• Record the rack elevations. Photograph the front and rear of every rack.
• Map the cabling and ports.
• Design the layout at the new end and get it agreed.
• Check power, cooling and floor loading there.
• Order the network and cross-connects.
• Agree the shutdown and start-up order.
• Separate out anything that's being disposed of.

1 to 4 weeks before: prepare

• Issue the RAMS and get them approved. Our Server Move RAMS Template is a start.
• Send the building management packs for both sites.
• Confirm the loading bay and lift bookings.
• Arrange access lists and passes.
• Agree how devices, cables and rails will be labelled.
• Confirm packing, cases and anti-static protection.
• Take full backups, and test a restore.
• Line up your support and vendor contacts for the window.
• Agree a rollback plan.
• Hold a final go or no-go meeting.

Shutdown and de-rack

• Change freeze in place and the final backup checked.
• Shut down in the agreed order, by authorised staff.
• Label each device with where it came from and where it's going.
• Label cables at both ends before you pull them.
• Bag rails, brackets and screws, and label them with their device.
• Check serial numbers against the list as each device comes out.
• Pack with anti-static and impact protection.

Transport

• Check the load against the asset list before the doors close.
• Secure the load against movement.
• Lock the vehicle, and record the seal reference if you use one.
• Share driver and crew details with the destination.
• Confirm the arrival time.

Re-rack and reconnect

• Check the count and serials on arrival.
• Look at the packaging for damage before you open it.
• Install to the agreed layout, heaviest kit lowest.
• Connect power to the planned feeds.
• Patch to the port map and tidy the cabling.
• Power up in the agreed order, by authorised staff.

Test and hand over

• Check the hardware health on every device.
• Check network connectivity.
• Have service owners test their services.
• Check monitoring and alerts.
• Reconcile the asset list and log exceptions.
• Clear the packaging and waste.
• Send redundant kit for certified disposal, if it's in scope.
• Get sign-off, and note what you'd do differently.

• No accurate asset list before the move. You find out what's missing when it doesn't turn up.
• Cables pulled without labels.
• Rails and screws packed away from their device.
• Backups taken but never test-restored.
• Power or cooling at the new end not checked.
• No agreed start-up order, and no rollback plan.
• A move window too tight to leave time for testing.

Most server move problems are planning problems. They're much easier to fix eight weeks ahead than at two in the morning.