Observability A–Z

A notification that a system health threshold has been crossed.

Software tooling that measures and tracks application performance in real time.

Amazon's cloud platform — the world's leading cloud infrastructure provider.

Microsoft's cloud platform for building, deploying, and managing applications.

Source code compiled into executable software artifacts, tracked by build number.

AWS service that logs all API activity across your AWS infrastructure.

AWS's monitoring and observability service for resources and applications.

The number of unique label combinations in a metric — high cardinality is a common scaling challenge.

A middleware pipeline that receives, processes, and exports telemetry from many sources to many backends.

Unified storage for logs, metrics, traces, and events — enabling deep correlation.

Practices combining development and operations to accelerate reliable software delivery.

Reducing data points in a time series to lower storage cost without losing insight.

A change event such as a pod restart, deployment, or config flag toggle.

A component that collects and transforms metrics from a system into a standard format.

A stacked visualisation that shows where time is spent across a call tree in a trace or profile.

A connection that brings telemetry from an external system into your observability platform.

The leading open-source container orchestration platform for deploying and scaling applications.

A recorded line of application output — structured or unstructured — for debugging and auditing.

The time elapsed between a request being made and a response being received.

Numeric measurements of system state over time — one of the three core observability signals.

Mean Time to Recover — the average time to restore a system after a failure.

Numeronym shorthand for 'observability' — 11 letters between the 'o' and 'y'.

The ability to understand system behaviour in production from its external telemetry outputs.

The standard OpenTelemetry protocol for transmitting telemetry between systems.

Prometheus's functional query language for selecting and aggregating time-series metrics.

The leading open-source monitoring system and time-series database for metrics.

Real User Monitoring — tracking performance as experienced by actual users in their browser.

A quantitative reliability target that defines acceptable service performance for users.

A single unit of work within a distributed trace, recording timing, name, and parent relationship.

Collecting a representative subset of traces to control volume and cost without losing insight.

A formal contract with customers defining the minimum reliability level a service must meet.

A Service Level Indicator — the specific metric used to measure whether an SLO is being met.

The automated collection and transmission of data from remote systems for monitoring and analysis.

A group of spans that together represent a complete request journey across services.