Service Level Agreement (SLA) is an agreement outlining the detailed characteristics of the service, the rights and obligations of the customer and the provider, as well as the availability level of the target service.
Among other things, the SLA sets forth:
- coordinated uptime (for example, “24×7” implies round-the-clock uptime, seven days a week);
- coordinated support time (for example, “8×5” implies “10:00-18:00, from Monday up to Friday”);
- response time.
The key point of the SLA is the actual availability, which specifies the percentage of the service availability for the pay period. It is calculated by the formula: (coordinated uptime per period – downtime per period)/coordinated uptime) × 100%. For example, the availability of 99.95% implies that the total downtime should not exceed 22 minutes per month. The rate can be calculated both for the infrastructure or network as a whole, and for individual utilities, depending on the type of service.
The SLA also specifies the level of disk subsystem performance (IOPS, latency), allocated time for technical maintenance (scheduled or urgent), compensation amounts.