Data Centre Decommissioning Checklist: A Phase-by-Phase Guide

• Written by: Simplify Networks
• November 7, 2026

Decommissioning gets far less attention than migration, yet it carries just as much risk. Once workloads have moved, the remaining hardware still holds data, still sits under lease and still has cabling, power and contracts attached to it. A clear data centre decommissioning checklist keeps the exit controlled, auditable and on time.

This checklist is written for service providers and project managers who run decommissioning projects for end customers and need onsite engineers to carry out the physical work. It assumes the migration itself is complete or close to it. If you are still planning the move, start with our data centre migration checklist and come back here for the exit.

Phase 1: Plan and confirm scope

Most decommissioning problems come from assumptions made before anyone visits the hall. Lock these down first:

1. Confirm every workload has moved and has been running in its new home for an agreed period. Get written sign-off from the application owners.
2. Build a full asset register for the racks in scope: make, model, serial number, asset tag, rack and U position. Reconcile it against the CMDB and the finance register.
3. Identify data-bearing devices. Servers, storage arrays, backup appliances, firewalls and switches with configuration all count. Mark each one with the required sanitisation method.
4. Agree the disposal route for each item: redeploy, return to lessor, resell or recycle through an approved disposal partner.
5. Check contracts: colocation notice periods, cross-connect and circuit cancellation lead times, support renewals and lease return conditions.
6. Book access early. Colocation providers in London, Frankfurt, Amsterdam and Paris often need named engineers, inductions and loading bay slots booked days in advance.

Phase 2: Prepare the site

• Produce rack elevations showing what comes out and what stays.
• Label every device and cable to be removed so the engineer does not rely on memory or on drawings that may be out of date.
• Confirm power-down order with the customer’s network and infrastructure teams, especially where shared switches or PDUs serve kit that must stay live.
• Arrange packaging, pallets, asset cages and secure transport.
• Agree who signs the chain-of-custody forms on each side.

Phase 3: Data sanitisation

This is the step auditors care about most. Every data-bearing device needs a recorded outcome.

Choose the method per device

Options usually include software wiping to a recognised standard, cryptographic erase for self-encrypting drives, degaussing, or physical destruction by shredding. The customer’s security policy normally decides. Where drives cannot leave the building, plan for onsite destruction or drive removal and secure storage.

Record everything

Capture the drive serial, host asset, method, date, engineer name and result for each drive. Photograph removed drives against their asset labels. Collect certificates of sanitisation or destruction and attach them to the asset register.

Phase 4: Physical removal

1. Power down in the agreed order and confirm with the remote team before pulling power cables.
2. Remove patch and power cables back to the source, not just at the device end. Abandoned cable is a common reason for disputes with colocation providers.
3. De-rack equipment, check serials against the register and tag each item with its disposal route.
4. Pack and palletise, then complete chain-of-custody forms at handover to the courier or disposal partner.
5. Remove rails, blanking panels and any customer-owned PDUs if the contract requires an empty rack.

Engineers who already do rack and stack work are well suited to this phase, because reverse racking needs the same discipline with weight, rails and cable management. Our rack and stack checklist covers the forward direction.

Phase 5: Close out

• Cancel circuits and cross-connects only after the hardware they serve is confirmed off.
• Walk the space with the colocation provider and get written confirmation it has been returned in the agreed condition.
• Update the CMDB, finance register and any support contracts.
• Issue a closing report with photos, the reconciled asset register and all sanitisation certificates.

Decommissioning across several sites, from Manchester to Stockholm

Large exits often span several facilities at once, for example consolidating halls in Manchester, Dublin and Stockholm into one cloud region. Use one asset register format, one evidence standard and one reporting template across every site, and stagger the work so lessons from the first site improve the rest.

Simplify Networks provides data centre decommissioning engineers who work under your brand, following your process and evidence requirements. Jobs are coordinated from London, with a 24/7 service desk responding in under 30 minutes. See our coverage and dispatch and SLA options, or read how we work with field service partners.

Where we deliver

We support decommissioning and data centre work in the UK, Germany, the Netherlands, France, Ireland and Sweden, among many other countries.

Frequently asked questions

What is the most important step in data centre decommissioning?

Data sanitisation. Every data-bearing device needs a recorded sanitisation or destruction outcome, linked to its serial number and backed by a certificate, so the customer can prove no data left the building unprotected.

Should cables be removed during decommissioning?

Usually yes. Most colocation contracts require the space to be returned clear, including patch, power and overhead cabling. Remove cables back to the source and confirm what the contract requires before work starts.

When should circuits and cross-connects be cancelled?

Only after the hardware they serve has been confirmed powered off and the customer has signed off. Check notice periods early, as some carriers and colocation providers need weeks of notice.

Can drives be destroyed on site?

Yes, where the customer’s security policy requires drives not to leave the building. Plan onsite shredding or degaussing in advance, or remove drives and hold them securely for a destruction partner.