Service level agreements (SLAs)

Service level agreements, or SLAs, address expectations, impacts, and consequences if agreed-upon reliability is not met. This is a contract that outlines what happens if a service meets—or misses—SLO targets. Consequences for missing SLOs are often business decisions or financial repercussions, such as:

The customer expects a given service to have a 0.05% maximum error rate daily, or they’ll receive a rebate. The customer expects only 10% of monthly requests to take longer than 500ms to complete, or they’ll be reimbursed for the compute overage. These stipulations are often related to business and financial relationships between the user and service. As such, SLAs are generally outside of the purview of Engineers. Instead, they’re often written by other company stakeholders before specific SLOs are quantified.

Engineering teams can derive their SLOs from the SLAs defined by other parts of the company, but SLO targets should generally be more strict than SLA targets