2.4Service Management

A system is not "finished once built"—it must be kept running stably day after day. How to respond when a failure occurs, how to prevent it from recurring, and how to manage the agreement with users: these are what service management addresses.

2.4.1ITIL, SLA, and SLM

• ITIL is a guideline compiling best practices for IT service management, widely referenced worldwide, forming the foundation for processes such as the service desk and incident management.
• An SLA (service level agreement) is a document in which the provider and the user agree in advance on the quality of the service to be provided (uptime, response time, etc.). SLM (service level management) is the activity of continuously managing whether the levels defined in the SLA are actually being maintained and improved.

2.4.2Incident/problem/change management and the service desk

• Incident management aims to restore normal service as quickly as possible when a service disruption or failure occurs (finding the root cause is not its goal). Problem management aims to investigate the root cause of incidents and prevent recurrence.
• Change management plans and executes changes to a system (configuration changes, new features, etc.) after evaluating the impact in advance and obtaining approval. It prevents uncontrolled changes from causing new failures.
• The service desk is the single point of contact for user inquiries and failure reports. Users simply contact this one place; internally, the service desk coordinates which team the request is routed to.

2.4.3Facility management

• Facility management maintains the physical facilities that support IT services, such as server rooms, power, and air conditioning. A UPS (uninterruptible power supply) keeps supplying power for a limited time during an outage, allowing the system to shut down safely (or switch to an emergency power source).
• Redundancy is the design idea of preparing multiple units, including spares, of servers or network equipment, so the service as a whole can continue even if part of it fails.

Trace an incident where users cannot log into an in-house core system. The user first contacts the service desk, which gathers symptoms and routes the case to the responsible team. The top priority here is not fully explaining the cause but simply getting the service usable again as quickly as possible, so a stopgap fix—say, restarting the authentication server—is applied first as part of incident management. Once login works again, the incident itself is considered resolved. But if the same failure recurs the following week, the root cause needs to be found. This is where problem management takes over: detailed log analysis might reveal the root cause as "the authentication server's memory runs out under a specific condition." Addressing this root cause requires a configuration change or software update on the authentication server, but applying it straight to production risks introducing new problems, so it goes through change management—evaluating the impact in advance and obtaining approval before applying it in a planned way. To continuously measure the quality of this whole operation, the provider and users agree in advance on an SLA (e.g. "99.9% or higher uptime during business hours"), and SLM continuously checks and improves whether actual performance meets that level. Furthermore, facility management measures such as a UPS that temporarily secures power so the server can shut down safely during an outage, or redundancy, keeping a spare server ready to take over if one fails, form the foundation that lowers how often such failures occur in the first place.

2.4.4Section summary

• ITIL is a collection of IT service management best practices. An SLA is the agreed document on quality; SLM is the activity that continuously manages those levels
• Incident management aims for quick restoration; problem management aims for the root cause and preventing recurrence; change management requires prior impact assessment and approval
• The service desk is the single point of contact for users. Facility management (UPS, redundancy) is the foundation that reduces failures from occurring in the first place