SRE Basics: SLIs, SLOs, SLAs and the Four Golden Signals

GCP Study Hub
August 13, 2026

Google's Site Reliability Engineering (SRE) guidance provides two sets of terms worth knowing for the Cloud Digital Leader exam. The first is the four golden signals, the metrics SRE names for monitoring system health: latency, traffic, errors and saturation. The second is the set of service-level terms. A service-level indicator (SLI) measures how a system is performing, a service-level objective (SLO) sets a target for that performance, and a service-level agreement (SLA) is a contract that defines commitments between a provider and a customer.

The four golden signals

  • Latency measures how long requests take to complete.
  • Traffic shows the demand placed on the system.
  • Errors capture failed or incorrect requests.
  • Saturation shows how close the system is to running at full capacity, or to being overloaded.

Tracking all four together lets teams identify and diagnose problems quickly and understand how well their services are functioning. Each signal looks at a different aspect of a live service: how fast it responds, how much it is being asked to do, how often requests fail, and how close it is to its limits. Together they describe operational health, which is a narrower question than how a system is designed or how its infrastructure is planned.

Service-level indicators

A service-level indicator is a measurable metric that reflects how a system is performing. Latency, throughput and error rate are typical examples. SLIs are the foundation of reliability engineering because they provide the evidence, in the form of objective numbers, that gets compared against targets. Each individual metric is something an SLI can track, and the SLI is the broader concept of how such metrics are used to judge system health.

Service-level objectives

A service-level objective is a performance goal or target. An SLO does not produce measurements of its own. It relies on SLIs to supply the numbers that show whether the goal is being met. The SLI reports how the system is performing, and the SLO states the level of performance the team is aiming for.

Service-level agreements

A service-level agreement is a contract between a provider and a customer that defines commitments. It is a contractual promise rather than a measurement.

Keeping the terms straight

A simple way to separate the three is by what each one does. The indicator measures, the objective sets the target and the agreement records the commitment.

Latency appears in both halves of this topic, once as a golden signal and once as an example of what an SLI might track. The terms still answer different questions. The golden signals name the dimensions of service health to monitor, while SLIs, SLOs and SLAs describe how measurements, targets and commitments relate to one another. Holding that distinction makes Cloud Digital Leader scenarios on reliability easier to work through.

Our Cloud Digital Leader practice exams cover SRE concepts such as the four golden signals and SLIs alongside Cloud Monitoring and Cloud Logging, with detailed explanations that drill these distinctions.

Get tips and updates from GCP Study Hub

arrow