Guidance

Service management good practice

Published 7 January 2015

IT service management

All service delivery organisations should use an IT Service Management (ITSM) framework, such as the IT Infrastructure Library (ITIL). This document uses ITIL language to set out what PSN organisations should expect of one another.

Any customer with a service-related matter must contact their service provider(s) in the first instance. Service providers should manage service-related matters such as events, incidents, problems, changes, releases and capacity issues in accordance with their Service Level Agreements (SLAs).

Incident management

An incident is an unplanned service event that impacts more than one organisation. At least one organisation needs to take action to rectify the incident in order to resume normal service to other impacted organisations.

Service providers normally only investigate incidents that are raised internally or by their own customers. Service providers are responsible for managing incidents within their own service architecture, and either confirming that the fault is not theirs or identifying the root cause and fixing it.

Problem management

A problem arises if an incident cannot be fixed immediately, or it is likely to recur and will require fixes whenever it occurs.

Service providers normally only investigate problems that are raised internally or by their own customers. Service providers are responsible for managing problems within their own service architecture, and either confirming that the fault is not theirs or identifying the root cause and fixing it.

Change management

A change is a planned event that may have a short term impact on more than one organisation, but ultimately will modify, fix or improve a service that is being delivered. Changes include releases.

Service providers normally only get involved in changes raised by their own supply chain or their customers. Each change will normally have:

  • a change authority - who authorises the change
  • a change owner - who executes the change
  • a change window - a period of time during which the change will occur, often out-of-hours for critical services
  • a back-out plan - should the change not be successful

When there is a need for an emergency change, the change owner may proceed without authorisation, but should do everything possible to inform those impacted.

Capacity management

Network capacity management is the planning and forecasting of bandwidth needs and scheduling improvements to meet these needs.

Service providers normally have their own capacity plans. These are driven by their own forecasts, the commitments that they have made to their customers, and their budgets and risk appetites. Providers should also have the means to react quickly to sudden changes in demand.