Data Centre Decommissioning

Decommissioning is a controlled process, not a clear-out

Stripping a rack is straightforward. Doing it so that every drive is accounted for, every serial is recorded, and nothing walks off between the cabinet and the vehicle is the actual work. Server and storage estates concentrate more data per kilogram than anything else we handle, and the asset values are high enough to be attractive.

Our process is built around that: documented chain of custody from your floor to our facility on every job and, on our Asset Reporting service, every asset logged individually with serialised destruction records at the end.

What a decommission covers

• Servers, storage arrays and networking hardware, removed from the rack and collected in our own vehicles.
• Drives handled separately from chassis where you need them removed and destroyed on their own.
• Data sanitisation to NIST 800-88 Rev 2 Purge, or physical destruction where a drive cannot be verified. See secure data destruction.
• Serialised records for every asset and drive through Asset Reporting, ready for your asset register and auditors.
• Valuation of enterprise hardware before anything is written off, through technology buyback.
• Recycling of whatever cannot be reused, zero to landfill, through licensed WEEE partners.

SSDs at scale

Flash storage changes the calculation. Degaussing does nothing to an SSD (our guide to NIST 800-88 explains why), so at data centre scale the real decision is between verified erasure and shredding.

Our approach: sanitise to NIST 800-88 Rev 2 Purge where the drive supports it and can be verified, and physically shred where it cannot. At data centre volumes the split matters commercially, because verified-wiped enterprise SSDs retain real resale value while shredded ones do not. We will tell you which is which rather than defaulting to whichever suits us.

Site realities

Data centre and comms room work has constraints that office collections do not:

• Access control. Pre-authorisation, photo ID, escort requirements and sometimes per-visit approval. Arrange it before the day.
• Floor loading and lifts. A full rack is heavy. Goods lift capacity and booking windows are frequently the binding constraint on how fast a job can go.
• Live adjacency. Most decommissions happen next to equipment that is very much still running. That dictates pace and method.
• Out-of-hours. Often the only workable window. Possible with notice.
• Structured cabling. Frequently forgotten in the plan, and a surprising amount of labour. Decide upfront whether it is in scope.

Residual value in enterprise hardware

Enterprise equipment holds value differently from office kit. Recent CPUs, populated memory, enterprise SSDs, GPUs and networking hardware all have active secondary markets, and a decommission of any size is worth valuing before it is written off. It is not unusual for the recoverable value of a modern rack to cover the cost of the decommissioning work.

The main things that destroy value: missing rails and caddies, asset engraving, and leaving equipment in a store room for two years after it came out of service.

Managed service provider or hosting company? If the hardware belongs to your clients, see IT disposal for data centres and MSPs for how we work under your name.