Guide
Data center decommissioning checklist.
Seven phases, the items that belong in each, and the mistakes that cost the most. Written from live enterprise and colocation projects — free to read, free to share, no form required.
Use this before you book the first access window
Decommissioning goes wrong in predictable places: scope that was never written down, inventories captured at pallet level instead of serial level, destruction evidence that cannot be tied to a device, and billing that keeps running after the racks are gone.
This checklist follows the sequence we use on enterprise and colocation projects. Work through it with your IT, security, facilities and finance owners in the room. Anything you cannot answer in Phase 1 becomes a change order later.
Want it as a document? Print this page to PDF, or ask us to send the formatted version with a scope template attached.
What the checklist covers
- 01Scope and stakeholder sign-off
- 02Asset inventory and data-bearing device identification
- 03Data destruction method and evidence standard
- 04Physical decommissioning and de-racking
- 05Packaging, transport and chain of custody
- 06Resale recovery and responsible recycling
- 07Reporting, audit closeout and contract release
Phase 1
Scope and stakeholder sign-off
Most decommissioning overruns are scoping failures, not execution failures. Fix the boundary of the project before anyone touches a rack.
- Confirm which rooms, cages, rows and rack positions are in scope, by label — not by description.
- List every system that must stay live through the project, including out-of-band management, cameras, door controllers and building automation.
- Identify circuits, cross-connects and carrier services to be terminated, with notice periods for each provider.
- Name a single decision owner for IT, security, facilities, finance and (if applicable) the landlord or colocation operator.
- Agree the target end date and the date the space must be returned in contractual condition.
- Record the data classification of the environment so destruction requirements are set before equipment moves.
- Decide up front whether resale recovery is in scope — it changes handling, packaging and timing.
Costliest mistake
Skipping the landlord or colocation operator early. Access windows, insurance certificates, floor-loading limits and elevator reservations are usually the longest lead items in the whole project.
Phase 2
Asset inventory and data-bearing device identification
Your inventory is the backbone of both the audit file and the resale value. Capture it at serial level, before disconnection.
- Capture manufacturer, model, serial number, rack and U-position for every asset.
- Reconcile the physical count against the CMDB or asset register and document every variance.
- Flag every data-bearing device: HDDs, SSDs, NVMe, tape, flash modules, RAID controllers with cache, and network devices holding configurations.
- Record drive counts per chassis — that number is what the destruction certificate must match.
- Photograph rack faces and rear cabling before removal, for restoration reference and dispute avoidance.
- Tag anything leased, financed or under a manufacturer return obligation so it is not destroyed by mistake.
- Note embedded storage in printers, appliances, KVMs, UPS management cards and PDUs — commonly missed.
Costliest mistake
Pallet-level or quantity-only inventories. An auditor asking 'prove this specific serial was destroyed' cannot be answered with a count, and the gap is usually found months later.
Phase 3
Data destruction method and evidence standard
Choose the method by media type and policy, then define what proof you will hold before the first device leaves the room.
- Match method to media: overwrite-based erasure for reusable drives, degaussing for magnetic media, crushing or shredding where reuse is not permitted.
- Confirm that solid-state media is not treated as magnetic media — degaussing does not sanitize SSD or NVMe.
- Define the evidence standard: serialized erasure reports, serialized destruction certificates, or both.
- Decide whether destruction happens before transport or in a secured, monitored warehouse under sealed chain of custody.
- Require tamper-evident seals with recorded seal numbers on every container of data-bearing media.
- Specify retention period and format for destruction records so they survive staff turnover.
- Confirm who is permitted to witness destruction and whether video evidence is required.
Costliest mistake
Accepting a single blanket certificate for an entire project. Blanket certificates cannot be tied to serials, which is exactly what a regulator or customer audit asks for.
Phase 4
Physical decommissioning and de-racking
Sequence the floor work so the room stays safe, powered correctly and reversible if something in scope turns out to still be live.
- Verify each system is out of service with the application owner before power-down — in writing.
- Power down in dependency order, then label and cap de-energized circuits.
- Remove and coil structured cabling and fibre; separate copper for recovery value.
- De-rack from the top down, keeping rails, shelves, PDUs and cable management with their cabinet.
- Keep chassis, caddies and drive trays together if assets are going to resale.
- Protect raised floor, walls, doorframes and elevators; document any pre-existing damage.
- Segregate hazardous items: UPS batteries, fire-suppression components, ballasts and capacitors.
- Leave the space in the contractual return condition — swept, cleared and with abandoned material removed.
Costliest mistake
Pulling cabling before the circuit map is confirmed. Once labels are gone, proving what was terminated becomes guesswork and can stall the contract closeout.
Phase 5
Packaging, transport and chain of custody
Custody is either continuous and documented, or it is a gap in your audit file. There is no middle state.
- Palletize and shrink-wrap by asset class; use anti-static packaging for boards and components.
- Seal data-bearing media in dedicated containers with recorded seal numbers.
- Issue a manifest per pallet and per container that ties back to the serialized inventory.
- Use vetted carriers with signature capture at pickup and at receipt.
- Confirm insurance coverage in transit, and who carries the risk at each handoff.
- For cross-border movement, prepare commercial documentation, valuation and any required export or import paperwork early.
- Reconcile received counts against shipped counts on arrival, and investigate any variance before processing.
Costliest mistake
Unsealed mixed pallets containing drives. Once media is mixed into general equipment, per-serial destruction evidence becomes impossible to reconstruct.
Phase 6
Resale recovery and responsible recycling
Recovery value is highest when equipment is tested and remarketed quickly, and when everything that cannot be reused is documented downstream.
- Test and grade functional equipment; capture configuration so it can be quoted accurately.
- Separate high-demand items — current-generation CPUs, memory, GPUs, optics and enterprise drives — from commodity material.
- Agree the settlement model in advance: outright purchase, revenue share, or offset against project cost.
- Require a downstream statement for material that is recycled rather than reused.
- Confirm recycling is performed under a recognized standard with audited downstream vendors.
- Capture weights by material stream for sustainability and ESG reporting.
- Ask for diversion-from-landfill figures if your organization reports on circularity.
Costliest mistake
Storing equipment for months before remarketing. Enterprise hardware depreciates fast; delay usually costs more than the storage saved.
Phase 7
Reporting, audit closeout and contract release
The project is finished when the paperwork would survive an audit without anyone needing to remember what happened.
- Assemble the final package: serialized inventory, erasure and destruction certificates, weights, downstream statement and settlement summary.
- Reconcile the final package against the original inventory, line by line, and explain every exception.
- Retire assets in the CMDB and fixed-asset register, and notify finance for write-off or disposal accounting.
- Close carrier, cross-connect and maintenance contracts, and confirm billing has stopped.
- Obtain written space acceptance from the landlord or colocation operator.
- Store the audit file where security, compliance and finance can retrieve it without the project team.
- Run a short post-project review and capture lessons for the next site.
Costliest mistake
Leaving recurring billing open. Cross-connects, maintenance and colocation charges routinely continue for months after the hardware has gone.
Where we fit
Run the checklist yourself, or hand the whole project over.
Decommissioning
Planning, de-racking, packaging and reporting managed end to end.
Chain of custody
How custody, destruction and evidence are recorded at every stage.
Certifications
The standards our recovery and destruction operations are held to.
Planning a specific site? See regional decommissioning coverage across the United States and Canada, or explore all services.
FAQ
Decommissioning questions
Send the site, rack count and target date and a specialist will reply with a scope, schedule and recovery estimate — usually within one business day.
Let's solve what's next
Stay connected.
Tell us what you need and we will connect you with the right USACA Group team — recovery, semiconductor, hardware or global supply chain.