OpenTelemetry: From Discovery to Expertise

Chapter Metrics

1. What is an exemplar?

A measurement kept with the trace_id of the request that produced it A dashboard template A Collector configuration example An instrument type

2. What is 'Head Sampling'?

Sampling done at the start of the trace Sampling done on errors only Sampling done by the collector A compression technique

3. In the collector, what does the 'resourcedetection' processor do?

It deletes data It queries the environment (cloud, K8s) to add labels to outgoing data It checks the YAML is valid It detects outages

Chapter The OpenTelemetry Collector

4. Which of these is NOT a component of the OTel collector?

Receiver Processor Exporter Database

5. Which is an example of a 'Semantic Convention' for HTTP?

request.url http.url url.full (or http.target in older versions) web.path

6. Can you limit the sending rate to a backend (rate limiting)?

No Yes, through processors or exporter settings Only by restarting Only in Java

7. What is an 'Attribute' in a span?

An attached file A key-value pair providing context (e.g. http.status_code: 200) A domain name A password

8. Which of these languages does NOT yet have a stable OTel SDK?

Java Python Go All the languages listed have stable or mature SDKs

Chapter The OpenTelemetry Collector

9. Which section of the collector configuration file defines the data flows?

receivers processors exporters service.pipelines

Chapter The OpenTelemetry Collector

10. What is the difference between OpenTelemetry and Prometheus?

Prometheus is a standard, OTel is a tool OTel is a multi-signal collection standard, Prometheus is mainly a metrics system and a database They are identical OTel replaces Prometheus

11. When is manual instrumentation necessary?

For Java applications only To capture specific business events or details not covered by the agent Never, the agent is always enough Only for logs

12. What is the main advantage of Java auto-instrumentation?

It looks nicer It relies on bytecode manipulation at runtime, requiring no recompilation of the code It is free It does not need a collector

13. OpenTelemetry is the result of merging which two projects?

OpenTracing and OpenCensus Jaeger and Zipkin Prometheus and StatsD Logstash and Fluentd

14. What is a 'Span Link'?

A hyperlink A causal relationship between two spans that are not necessarily parent/child (e.g. asynchronous processing) A keyboard shortcut A network error

Chapter Deployment, Kubernetes and case studies

15. With the OpenTelemetry Operator, how do you instrument a Java application?

With the instrumentation.opentelemetry.io/inject-java annotation and an Instrumentation resource By rebuilding the image with the agent By adding a Jaeger sidecar It is not possible

Chapter Introduction and Architecture · Alloy: From Discovery to Expertise

16. What does 'Signal' mean in OTel terminology?

An alert A category of telemetry data (Trace, Metric, Log, Profile) A radio signal A CPU interrupt

Chapter Introduction and Architecture · Alloy: From Discovery to Expertise

17. What is the difference between a Counter and a Gauge?

The Gauge is faster A Counter can only go up, a Gauge can go up and down The Counter is for logs There is no difference

18. What is a 'Log Bridge API'?

A physical bridge An API that lets classic logging libraries send data to OTel A language translator A paid API

Chapter Introduction and Architecture · Alloy: From Discovery to Expertise

19. What are the three historical pillars of observability supported by OTel?

Logs, Metrics, Traces CPU, RAM, Disk Alerts, Dashboards, Reports Events, Profiles, Debug

20. What is the impact of OTel instrumentation on CPU performance?

Very high (> 50%) Low (usually < 5%) if configured properly None Impossible to measure