Capital Planning & Project Delivery

Data centre operations, defined.

The terms that come up when you run, maintain and invest in critical infrastructure – from CMMS and PPM to MTBF, permits and SLAs – each in a sentence or two.

Operations leaders reviewing a strategy paper

30 terms

The working language of
critical operations.

Plain definitions of the maintenance, reliability, service and capital terms used across data-centre operations, with a link to the Clarisync page that covers each one.

A

Asset management
The coordinated work of keeping physical assets – plant, power, cooling and building systems – known, maintained and fit for purpose across their life. In a data centre it covers the asset register, maintenance, condition and the evidence that work was done.
Related: Asset Management

C

Capital expenditure (CAPEX)
Spending on new assets or on work that extends an existing asset's life, such as replacing chillers or upgrading UPS systems. It is planned and approved separately from day-to-day running costs.
Related: Capital Planning
Capital planning
Deciding which capital works to fund, when, and in what order, based on asset condition, risk, service impact and budget. The output is a prioritised, costed programme over several years.
Related: Capital Planning
Change Advisory Board (CAB)
The group that reviews and approves proposed changes to live systems or infrastructure, weighing risk, timing and back-out plans before work goes ahead.
Related: Change Management
Change management
The controlled process for proposing, assessing, approving, scheduling and recording changes to a live environment, so that changes do not cause unplanned outages.
Related: Change Management
Competency management
Recording and checking that each person – employee or contractor – holds the training, authorisations and experience a task requires before they are assigned to it.
Related: Workforce & Competency
Computerised maintenance management system (CMMS)
Software that records assets and plans, schedules and tracks maintenance work orders. A CMMS focuses on maintenance execution; an EAM system covers the wider asset lifecycle.
Related: Asset Management
Condition-based maintenance
Maintenance carried out when measured condition – such as vibration, temperature or insulation readings – shows it is needed, rather than on a fixed calendar.
Corrective maintenance
Work done to restore an asset after a fault or failure has been found. It can be urgent (reactive) or planned for a later window.

E

Emergency operating procedure (EOP)
A pre-approved, step-by-step response for a specific failure or emergency, such as loss of utility power, so staff act correctly under pressure.
Related: Risk & Reliability
Enterprise asset management (EAM)
Managing assets across their whole lifecycle – planning, acquisition, operation, maintenance and disposal – often across many sites. It is broader than a CMMS, which concentrates on maintenance work.
Related: Asset Management

H

Human-in-the-loop
An AI design in which a person reviews and approves consequential outputs before they take effect. The AI proposes; an accountable person decides.
Related: AI Transparency
Hyperscaler
A very large cloud or internet company that operates data centres at global scale, often building and running its own campuses.
Related: Who We Serve

I

Incident management
Restoring normal service as quickly as possible after an unplanned interruption or degradation, and recording what happened and what was done.
Related: Incident & Problem

M

Mean time between failures (MTBF)
The average operating time between one failure of a repairable asset and the next. Higher MTBF indicates more reliable equipment.
Related: Risk & Reliability
Mean time to repair (MTTR)
The average time taken to repair a failed asset and return it to service. Lower MTTR shortens the impact of each failure.
Related: Risk & Reliability
Method of procedure (MOP)
A detailed, step-by-step plan for a specific piece of work on live infrastructure, including checks, hold points and how to back out if something goes wrong.
Related: Risk & Reliability
Monthly operational review (MOR)
A regular report and meeting between a service provider and its client covering performance, incidents, changes, maintenance and SLA results for the month.
Related: Command

N

N+1 redundancy
A design with one more component than is needed to carry the load (N), so a single unit can fail or be taken out for maintenance without losing capacity.

O

Operational expenditure (OPEX)
The ongoing cost of running a facility – maintenance, energy, staff, contracts and consumables – as opposed to capital spending on new or life-extending assets.
Related: Operational Finance
Owner-operator
An organisation that both owns its critical facilities and operates them itself, rather than relying on a third party to run them.
Related: Who We Serve

P

Permit to work
A formal written authorisation for hazardous or high-risk work, stating what may be done, by whom, where, when and under which safety controls.
Related: Workforce & Competency
Planned preventive maintenance (PPM)
Maintenance scheduled at set intervals or usage levels to reduce the chance of failure, such as servicing generators or replacing filters before they degrade.
Related: Asset Management
Power usage effectiveness (PUE)
A measure of data-centre energy efficiency: total facility energy divided by the energy used by IT equipment. A value of 1.0 would mean every unit of energy reaches the IT load.
Problem management
Finding and removing the root cause of recurring or significant incidents, so the same failure does not happen again.
Related: Incident & Problem

Q

Quarterly business review (QBR)
A quarterly meeting between a provider and its client that reviews performance against the contract, trends, risks and plans for the next period.
Related: Command

R

Reliability-centred maintenance (RCM)
A method for choosing the right maintenance strategy for each asset based on how it can fail and what the consequences of failure would be.
Related: Risk & Reliability
Retrieval-augmented generation (RAG)
An AI technique that retrieves relevant records at the moment a question is asked and uses them to ground the answer, instead of relying only on what a model learned in training.
Related: AI Transparency

S

Service level agreement (SLA)
The contractual commitment between a provider and its customer on measurable service levels, such as response times, resolution times and availability.
Related: Service Delivery
Standard operating procedure (SOP)
A documented, approved way of carrying out a routine task, so that it is done the same, safe way every time.
Related: Risk & Reliability

See the terms in a working operation.

A demonstration on your own assets, people and work shows how these ideas connect in one operating model.