The main New Relic glossary contains the list of terminology commonly encountered by New Relic users. This glossary containts the list of terminology particularly important for developers trying to understand how the Telemetry SDKs work.
A collection of measurements or other data at a remote system or inaccessible points. Telemetry in the Telemetry SDKs context currently refers to metrics or traces.
The harvester is a higher level construct that exposes methods to record telemetry, batching the telemetry, aggregating the telemetry, and sending the data to New Relic at an interval.
A batch is a collection of telemetry data of the same type, e.g. a batch of metrics, a batch of spans. Telemetry SDKs sends batches of telemetry data to New Relic. A batch can also have attributes that are common for all contained telemetry data.
A metric, at a conceptual level, is a measurement. The measurement is identified by a name, type, and key/value attributes. The New Relic Metric API accepts batches of metrics sent by the Telemetry SDKs.
Multiple Metric data can be combined into a single metric by calculating the combined measurements. The type of calculation varies by metric type.
Represents the value of something at a specific moment in time. The value can go up or down over time. No aggregation occurs with the gauge metric type.
Measures the occurrences of an event over a time interval. When count metrics are aggregated, the new resulting count metric has a count that is the sum of the input count metrics.
Used for reporting the count, average, sum, min, and max values of discrete events over a time interval. When summary metrics are aggregated, each summary value is recalculated.
A trace or a distributed trace is a causal chain of events between different components. It is composed of a collection of spans sharing the same traceId. The New Relic Trace API accepts batches of spans sent by the Telemetry SDKs.
A span is a component of a trace. For example, a database call or internal method call. See also trace.