ITIL Service Delivery: An Overview
Where IT service support is concerned with the day-to-day handling of incidents and requests, ITIL service delivery is the longer-horizon discipline of planning and maintaining the IT services themselves. Its core processes include service-level management (defining and tracking the agreed performance standards for a service), capacity management (ensuring IT infrastructure can meet current and future demand without over-provisioning), availability management (designing services to meet agreed uptime targets), and IT service continuity management (planning for recovery after a major disruption).
Financial management for IT services is often grouped alongside these disciplines as well, covering the budgeting, accounting, and charging models that let an organization understand the true cost of the services it delivers. Taken together, these processes are what separate an IT department that reacts to problems as they occur from one that plans service capacity and continuity ahead of demand.
Service-level management is usually where the other disciplines get anchored in practice. A service-level agreement (SLA) sets the performance targets a provider commits to for the business it serves; an operational-level agreement (OLA) sets the matching internal commitments between IT teams that have to cooperate to hit that SLA; and an underpinning contract (UC) does the same with an external supplier whose own performance feeds into the service. Capacity management is typically split the same way — business capacity management forecasts future demand from planned business activity, service capacity management monitors the performance of live services against their targets, and component capacity management tracks the individual servers, network links, and storage that those services run on.
IT service continuity management is distinct from ordinary backup and recovery: it covers the recovery of entire services after a major disruption — a data center outage, a widespread infrastructure failure — against an agreed recovery time objective and recovery point objective, not just the restoration of individual files. In ITIL's own framing, a continuity plan that has never been tested is not a plan at all — so the discipline includes periodic invocation testing alongside the documentation itself.