Skip to main content
Data Center Exit
A data hall being decommissioned: two rack rows, half emptied, coiled cable and pallets in the cold aisle

Decommissioning · ITAD · Destruction

A data center exit is a chain of custody problem before it is a logistics problem

An empty hall looks the same whoever emptied it. What differs is whether you can answer one question two years later in an audit: what happened to asset number such-and-such.

The certificate pack is the product

Serial-number-level records, the method per device, the destination of every asset, reconciled against the register you supplied. Everything before it was logistics.

The destruction policy is yours

A vendor funded by resale is better off erasing than shredding. That is legitimate, and it must never decide which of your assets get destroyed.

Access decides the schedule

Dock, elevator, route and out-of-hours rules move a project more than rack count does. They are settled in the site walk, not on the phone.

Cutaway of an exit sequence: rack row, floor void, equipment path to staging, destruction station

The devices nobody lists

A switch holds credentials, and it is not on the rack elevation

Management controllers, switch and firewall configuration, battery-backed cache on storage controllers, printers with internal drives, tape in a cupboard nobody has opened for years. None of it appears in an equipment list, and all of it holds something.

The full list, and what each one needs

Four questions, then a call with the floor plan open

Rack count, drive count and media types decide the price, and they come off the asset register rather than out of memory. We ask for the register on the call, not in a form.

Data Center Exit decommissions data halls: asset inventory, de-cabling, rack removal, on-site or witnessed data destruction, resale and recycling, and the certificate trail that proves each of them happened.

  • A blank certificate pack shown before we discuss price.
  • Certification numbers and expiry dates, verifiable with the certifying bodies.
  • Every subcontracted stage named, because each handover is a custody point.

Four questions

Request a project review

Rack count, drive count and media types decide the price, and they come off the asset register rather than out of memory.

To withdraw consent, reply STOP to any text or say so on the phone. The number is removed the same day. How the details are handled is set out in the privacy policy.

Chain of custodyCertificate of destructionAsset registerSerial-level reconciliationData sanitisationDegauss or shredResale recoveryCage decommissioningPower whip removalExit date

Where custody actually breaks

A list that stops matching the floor cannot be repaired later

  1. A single fully populated server rack in a data hall with cabling dressed down one side

    In production

    Racks working, cabling dressed, nothing to worry about. This is the only moment when an asset register can still be reconciled against what is physically in the cage, and it is the moment it almost never gets done.

  2. The same rack half emptied, its upper half stripped to bare rails with cables cut back and coiled

    Mid-exit

    Half the rack is gone and half is still live. This is where custody breaks: a drive pulled and set aside is no longer on anyone's list, and a list that stops matching the floor cannot produce a defensible certificate later.

  3. The same rack position completely empty, bare uprights with a floor tile lifted beside it

    Handed back

    Bare uprights, whips capped, floor swept. What proves the exit is not the empty cage but the paperwork behind it: every serial accounted for, destroyed or resold, each with a document naming it.

What an exit looks like

Five stages, five different risks

A wide cold aisle in a data hall with rows of closed racks on a raised floor and containment panels overhead

Before

Live data hall

A caged colocation suite with its mesh door open, half the racks removed and floor tiles lifted at the far end

Colocation

Caged suite

Pallets of decommissioned rack-mount chassis shrink wrapped and stacked on a loading dock with numbered tags

Logistics

Outbound pallets

Loose hard drives laid out in an anti-static tray on a stainless bench beside a degausser cabinet

Sanitisation

Drives and degausser

An empty data hall after decommissioning, bare raised floor with cable cutouts and power whips coiled and capped

After

Empty hall

The document, not the promise

Sanitisation without a serial-level record is a claim. What survives an audit, a regulator or a buyer’s diligence is a document that names each device, states the method used on it, and can be tied back to the register the cage was emptied against. That is built during the exit, not written afterwards.

What the paperwork has to say

Three questions that come up every time

Who performs the work?

We do. Data Center Exit decommissions data halls: asset inventory, de-cabling, rack removal, on-site or witnessed data destruction, resale and recycling, and the certificate trail that proves each of them happened. Where a stage is subcontracted we say so and name the facility, because every handover is a point in the chain of custody.

Destruction or erasure?

Both are defensible; which one applies to a given asset is a data classification decision and it is yours. The trade-off, including the media types that erase differently, is in the advice section.

Why is no price shown here?

Because access and drive count move a quote more than rack count does, and several factors move it in opposite directions. What a quote is made of is on the cost page.

All frequently asked questions