A colocation data centre is a facility built and operated to keep other organisations’ equipment running. Power architecture, cooling design, connectivity, security controls and the operational team behind them are all part of what you are buying.
Comparisons between UK colocation providers get useful when they get specific. Whether a facility describes itself as resilient tells you very little. How that resilience is designed, what evidence sits behind its security claims and who responds when something needs attention are questions that can actually be compared.
At Datum, the facility and the service around it are part of the same proposition. Farnborough and Manchester provide secure, carrier-neutral UK colocation, with service managers, operations teams and in-house engineers around the infrastructure.
Colocation space starts with the rack unit, or U: 44.45 mm of vertical space in a standard 19-inch rack. A full rack is commonly 42U, though larger cabinets are used.
Small deployments take part of a shared cabinet. A full rack gives one organisation the whole cabinet. Larger estates move into locked cages or private suites, where space and power are committed on a broader scale.
Having rack space for your equipment matters, but without sufficient power to support it, that space has little value.
Power density is quoted in kW per rack. Confirm the maximum supported density, whether that figure is a hard per-rack limit or an average across the hall, and how more power can be added as the estate changes. The gap between an average and a limit is where deployments can come unstuck.
Then ask whether the rack receives A and B feeds, how those paths are physically separated and what redundancy model sits behind them.
Cooling deserves the same treatment, and the vocabulary is shared. N means enough capacity for the current load, N+1 adds a spare, and 2N duplicates the system. Power and cooling can run different architectures, so ask about each one.
Hot and cold aisle containment controls airflow around equipment. The question worth asking is what happens to the environment when part of the cooling system fails.
-
Connectivity and carrier choice
Connectivity is one of the main reasons organisations move to colocation in the first place.
A carrier-neutral colocation facility lets you choose between network providers instead of accepting whichever one comes with the building. Resilience becomes easier to design, and changing carrier later becomes less of a project.
Ask for the current carrier list, not just the headline number. Then ask how the fibre enters the building. Two carriers sharing a duct still share a point of failure.
Cross-connects are the physical links between your rack and carrier equipment elsewhere in the facility. Internet access can come through IP transit or through peering at an internet exchange.
For hybrid strategies, private cloud connections can link colocated infrastructure directly into public cloud platforms without routing everything across the public internet.
-
Security, compliance and resilience
Physical security works in layers: perimeter controls, security staff, CCTV, controlled doors, biometric or card access, visitor procedures, cabinet locks and segregated cages. Which of those apply at the specific site, and how they are operated day to day, is the part worth asking about.
Certifications are evidence of a different kind. ISO 27001 is earned through ongoing surveillance audits, not a one-off assessment, and covers information security management specifically. PCI DSS applies where payment card data is involved, and accreditations such as NSI Gold or PASF speak to physical security.
Always ask for the scope. A company-wide certificate does not automatically mean every facility sits inside it.
Where personal data is involved, UK GDPR responsibilities follow it. If the provider is acting as a processor, for example by handling backups under a managed service, the contract needs to reflect that. Where data residency requirements apply, the physical location of the data centre is part of the answer.
The people around the facility matter as much as the plant inside it.
Remote hands gives you engineers who can carry out physical work when your own team is not on site. Service management adds deployment planning, reporting, regular reviews and operational visibility.
Check what support is included, what is chargeable and how out-of-hours work is defined. Those details can vary significantly between providers.
At Datum, this layer is deliberate. The facility gives you the secure, resilient environment. The engineers, service managers and operations teams help plan the move, deploy the estate and keep the work visible once it is live.