This set builds vocabulary for application performance monitoring and distributed tracing concepts.
0 / 5 completed
1 / 5
At standup, a dev references the tool tracking how a single request flows through multiple services, measuring latency at each hop. What is this capability called?
APM with distributed tracing tracks a single request as it flows across multiple services, recording latency and errors at each hop to reconstruct the full request path. This visibility is essential in microservice architectures where a single user action can touch many services. It goes beyond single-service logging to show cross-service behavior.
2 / 5
During a design review, the team wants to see which specific service in a request's path contributed the most latency. Which concept supports this?
A span represents a single unit of work, such as one service's handling of a request, within a larger distributed trace, and comparing span durations reveals which hop contributed the most latency. Aggregating and visualizing spans is core to root-causing slow requests in a multi-service system. This granularity is what distinguishes tracing from simple aggregate latency metrics.
3 / 5
In a code review, a dev sets up an alert that fires when a service's error rate crosses a defined threshold over a rolling window. What is this called?
A monitor (or alert) continuously evaluates a metric, like error rate, against a defined threshold and notifies the team when it's breached, enabling proactive detection rather than relying on someone noticing a dashboard manually. Well-tuned thresholds balance catching real problems against noisy false alarms. Monitors are a foundational building block of an observability setup.
4 / 5
An incident report shows a slow request was hard to diagnose because instrumentation was missing from one internal service in its path. What gap does this reveal?
If one service in a request's path lacks tracing instrumentation, the resulting trace has a blind spot, making it harder to pinpoint where latency or errors actually originated. Consistent instrumentation coverage across all services in a request path is necessary for tracing to be fully useful. This gap is a common finding when postmortems reveal a diagnosis took longer than expected.
5 / 5
During a PR review, a teammate asks how distributed tracing differs from simply aggregating each service's individual latency metrics separately. What is the distinction?
Aggregated per-service metrics show each service's overall performance in isolation, while distributed tracing connects the dots for one specific request across every service it touched, revealing the actual end-to-end path and where time was spent within that single flow. This request-level view is essential for diagnosing issues that only manifest in specific cross-service interactions. Both approaches are complementary parts of a full observability strategy.