A Service Level Agreement (SLA) is often considered the main contract with an ISP; however, the SLA is actually the technical aspect of the complete Service Level Contract (SLC), where the service levels are formally defined. The SLA often refers to contracted delivery times, values, parameters, and features that are negotiated, and it often describes, in plain language, what the organization is purchasing from the ISP.
General Considerations for SLAs
Many companies, vendors, and ISPs must provide SLCs to their partners or customers, and the SLA is a component of the SLC. The SLA defines specific parameters and performance measurements between devices (e.g., routers, servers, workstations, or other equipment). The SLC designates connectivity and performance levels that the ISP guarantees to the organization and that the organization guarantees to the end-users. An SLA is a common agreement of features such as the following:
- Services
- Priorities
- Responsibilities of the ISP and the organization
- Guarantees
- Warranties
SLAs might contain technical definitions, such as data rates, Mean Time To Failure (MTTF), Mean Time Between Failures (MTBF), or Mean Time To Repair (MTTR), which are predicted times to certain events in the system. All of these parameters depend on system failure and recovery definitions, as agreed to with the customer. SLAs can often be made up of smaller components known as Operating Level Agreements (OLAs). An OLA is often used internally in the organization and it describes what will be provided to the organization.
The SLA should contain measurable targets, committed information rates, and distinct levels of services defined in terms of minimum/maximum targets. The contract should also specify actions that should be taken in situations of non-compliance. The different levels of an SLA include the following:
- Customer-Based SLA: Defines the services offered to a particular customer.
- Service-Based SLA: Defines a particular service that can be offered to a customer or to multiple customers.
- Multilevel SLA: Includes multiple levels, each one of which has a different set of customers for the same services.
- Corporate-Level SLA
- Customer-Level SLA
- Service-Level SLA
The main resource regarding Cisco SLA concepts is the Cisco SLA portal at www.cisco.com/go/saa, where several white papers with information about implementing SLCs can be found:
- Service Level Management: Best Practices
- Deploying Service-Level Management in an Enterprise
- Service-Level Management: Defining and Monitoring Service Levels in the Enterprise
SLA Monitoring
From a technical standpoint, SLA monitoring can be achieved using tools like IP SLA, which is a service that allows monitoring, baselining, and optimization by generating many different types of synthetic traffic. IP SLA can be implemented on a variety of Cisco devices (IP SLA technical details were covered in the first chapter of this manual). SLA monitoring should take into consideration the metrics that define service baselines and that relate to the following:
- Ongoing diagnostics
- Comparison between different service levels and results
- Troubleshooting
- Optimization
SLA monitoring can be considered an early warning system in knowing whether there will be an outage or a service disruption from the ISP. SLA performance should be audited on a regular basis using SLA monitoring. SLA monitoring will also contribute to the continual improvement of the Information Technology Infrastructure Library (ITIL), which is a set of best practices related to IT service management.