observability

In software, observability describes how well you can understand internal behaviour from the data a system produces. This includes loggingGlossary · In briefloggingLogging records events while software is running. These records help developers follow a system's behaviour and investigate problems.Read more, metrics and traces. Together, these signals can reveal where a request slows down, which error occurred and which components were involved.

Which signals support observability?

Logs record events such as error messages or failed sign-ins. Metrics provide measurements over a period, such as response times, error rates or memory usage. Traces follow a request through an application's different components. Combining these sources provides more context than examining each separately.

Collecting large amounts of data does not automatically make a system understandable. Signals need sufficient context and must be possible to connect. A request ID can link an error message to its trace, user and releaseGlossary · In briefreleaseA release is an identifiable software version prepared to be made available to users. It brings together one or more checked changes.Read more, for example.

Observability and monitoring

MonitoringGlossary · In briefmonitoringMonitoring systematically tracks software, infrastructure or processes through measurements and checks. Deviations can trigger alerts so an administrator can investigate.Read more tracks predefined measurements and rules. An alert can report that response time has exceeded a configured threshold. This helps detect known deviations quickly.

Observability helps answer questions for which no dashboard or alert was configured in advance. Developers can investigate and combine available signals to discover why a deviation occurs. The concepts complement one another: monitoring reveals a problem, while observability helps explain the behaviour behind it.