What is Observability?

The property of how well the internal state of a system can be inferred from the data it emits. The data often named for it are metrics (numeric time series), logs (records of events) and traces (records of the path and timing of one request through several services). It aims not only at monitoring failures anticipated in advance but also at making it possible to investigate the causes of unanticipated problems. It is a different context from circuit observability in design for testability.

Japanese name: オブザーバビリティ

Observability appears in 1 practice question across 1 topic, including Cloud and Container Basics.

Field: Cloud and Container Basics

Try it: questions about Observability

In a system in which several services work together, you want to find out which services a single user request passed through, in what order, and how much time it spent in each. Among the data used for observability, which of the following is the most suitable?

AMetrics (numeric time series such as CPU utilization and request counts)
BLogs (records of events output by each service)
CTraces (records of the path of one request and the time of each segment)

Show answer and explanation

Answer: C. Traces (records of the path of one request and the time of each segment)

Observability is the degree to which the internal state of a system can be inferred from the data it emits, and metrics, logs and traces are often named as the data for it. In tracing (distributed tracing), an identifier shared by the request is passed between services, and the start and end times of the processing segments (spans) in each service are recorded to reconstruct the path of one request and the breakdown of its time. Metrics suit detecting overall trends and anomalies, and logs suit detailed investigation of individual events, but traces are the most suitable for following the flow of one request across services. In practice, the trace identifier is also written into the logs, and the three are combined to find causes.

Observability in the explanations

Observability is the degree to which the internal state of a system can be inferred from the data it emits, and metrics, logs and traces are often named as the data for it.
In a system in which several services work together, you wa…

Questions that mention Observability (1)

In a system in which several services work together, you want to find out which services a single user request passed through, in what order, and how much time it spent in each. Among the data used for observability, which of the following is the most suitable?

Books on Cloud and Container Basics Ad

Search Amazon for semiconductor Cloud and Container Basics

As an Amazon Associate I earn from qualifying purchases.

Back to the glossary