SLA for Incident Managemenent is about measuring time with timers to check that processes are executed in due time.

Service Level Agreement Management is directly implemented in the itil-common bundle with a few entities and are using Business calendars to measure time with hollidays time zones and business hours.

There are two main concepts for SLA Management:

1. Service Level Agreement Definition define what should happen (for example maximum duration for a process)
2. SLA Timers measure what actually happens (for example actual resolution time)

Service Level Agreement Definition

The SLA template (or definition) is the itilServiceLevelAgreement entity.

You can have the list of SLA with the menu IncidentManagement / SLA:

On each SLA you can define:

• the name
• if the SLA measure times (time recording)
• the Business Calendar used to measure business hours (hollidays, timezone, and business hours)
• the deadline duration: this is the total amount of time on which the process should be completed
• Start On / Stop On are imTicket properties. If they are filled on the incident, those properties are used to automatically start / stop the SLA timer. For example, you may want to measure resolution time and you define this time as the duration between submission time and closure time. Using the two corresponding properties will make sure that the status of the timer is set and the proper duration is calculated.
• Alerts: with alerts you can setup notifications to user, groups or calculated assignee (for example the group in charge of the ticket). You can also set the SLA level of the timer to be able to sort and filter incidents by SLA priority level or assign color code to alerts.
• Alert notification may have a reminder : as long as the alert is active, a reminder will be sent every xxx minutes for example.
• You can setup several alerts with various degrees of urgency, notify different groups

SLA Timers

SLA Timers are described by the itilSLATimer entity in the itil-common bundle. SLA Timers are components of imTicket (incident). That means that each timer is related and "owned" by an incident.

A timer has references to the SLA definition to handle its calculations and the incident.

It has also:

• a status: Started, Stopped, Paused
• a start time
• a deadline time = start time + maximum duration - pause time
• an end time: time when the timer is actually stopped
• pauses
• alerts
• breach duration = duration between deadline and end time when end time > deadline

Timers are managed by the application:

At the incident level (imTicket) there are up to 5 SLA Definitions.  Those SLA definitions (Applicable SLA 1 to 5) are used to track time on 5 different processes to manage incidents (for example resolution time, answer time, ...). Those 5 properties may be relabeled with a fragment on imTicket to adapt to your processes names.

Once those SLA definitions are filled (either manually or more realistically programmatically through scripting), the application will manage the Timers based on the SLA Definition.

So for example:

When a ticket is submitted, the "Submit" operation may lookup the appropriate SLAs definition based on priority and fill itilSLA1 for answer time, itilSLA2 for resolution time, etc...

Once the SLA definitions are set, the timers will be automatically created and managed.

You may also manually (programmatically) manage the timer with the following operations (on itilSLATimer):

• Start: fill the start time with the current date and the dead line
• Stop: fill the end time with the current date and calculate the actual duration and possible breach time
• Pause: void the deadline and add a pause
• Continue: resume the timer and recalculate the deadline taking in account the pauses
• Reset: void start, end time and deadline

Alerts

Alerts are defined in itilServiceLevelAlert (definition of the alert), as a component of itilServiceLevelAgreement. There may be several alerts on a SLA Definition.

Actual alerts (instances related to an incident) are managed in the itilSLATimerAlert entity.

There is a thread that pools the alerts based on trigger time (triggertime > now and status == active) and sends the notifications or updates the SLA level.