Lexicon
Telemetry
Telemetry is the data systems emit about their own behavior: logs, metrics, and traces. What each signal answers, and how the three connect in practice.
Definition
Telemetry is the data software and infrastructure emit about their own behavior. In observability it means three signals: logs, the record of what happened; metrics, the numbers that trend over time; and traces, the paths requests take through the system.
What it means in observability
Telemetry is the raw material everything else is made from, and each signal answers a different shape of question. Metrics are cheap and fast, ideal for is anything wrong. Traces explain where in a distributed flow the problem lives. Logs carry the detail of what exactly happened at that spot. Mature practices treat them as one connected dataset rather than three products, because real investigations hop between signals constantly: an alert on a metric, a trace to localize it, logs to explain it.
How it works in practice
Instrumentation produces telemetry inside applications and platforms, collectors and pipelines transport and shape it, and backends store and index each signal for its access pattern. What binds the signals together is shared context: resource attributes identify the emitting service and environment on every signal, and trace identifiers on log records connect logs to the requests that produced them. That shared context is what turns three data streams into one investigation surface.
Where it gets hard
Volume is the defining problem. Telemetry grows faster than the systems it describes, and AI era architectures accelerated the curve, so cost pressure pushes teams toward dropping, sampling, and shortening retention, each of which trades away coverage. Correlation is the other perennial gap: signals collected by different tools with different metadata never quite line up, and engineers pay the difference during incidents.
Where Tsuga fits
Tsuga treats logs, metrics, and traces as one correlated dataset with shared context, priced flat per GB and stored inside your own cloud account. The economics are designed so keeping your telemetry is the default, not the luxury.
Related terms
- AIOpsAIOps is the application of machine learning and artificial intelligence to IT operations: detecting anomalies, correlating and grouping alerts, suggesting root causes, and increasingly, powering assistants that investigate telemetry conversationally.
- Egress costsEgress costs are the fees cloud providers charge for data leaving their network, priced per gigabyte and varying by destination.
- InstrumentationInstrumentation is the code, runtime configuration, or platform setup that makes software emit telemetry.
- Log levelsA log level is the severity label attached to a log line: DEBUG, INFO, WARN, ERROR, and FATAL in the common scheme.
- Log managementLog management is the practice of collecting, processing, storing, and searching log data at scale: every event record your applications and infrastructure emit, made findable when someone needs it.
- MetricA metric is a named numeric measurement tracked over time, request counts, memory usage, response latency, carrying a type, a unit, and attributes that slice it into series.
- ObservabilityObservability is the ability to understand what is happening inside a system from the data it emits: its logs, metrics, and traces.
- Observability pipelineAn observability pipeline is the layer that moves telemetry from where it is produced to where it is stored, transforming it along the way: parsing, enriching, redacting, filtering, sampling, and routing to one or more destinations.
- TraceA trace is the end to end record of one request or workflow as it moves through a system, composed of all the spans that share a single trace ID.