Cloud Native Observability GCSA Practice Questions (Page 7)

Questions 31–35

1. 31application · medium

   A development team is building a new order-processing service. They want to ensure that when an order fails, they can quickly determine which step failed and why. Which combination of practices best supports this goal?

   • AUse structured logging with a consistent schema, add a unique order ID to every log, and create a trace span for each processing step.
   • BLog every event as free-form text and rely on full-text search to find errors.
   • CUse a single trace for the entire order process without sub-spans.
   • DRecord only high-level metrics like total orders processed and error count.
   Reveal Solution Discussion Select an answer first
2. 32application · medium

   A startup uses Kubernetes and wants to adopt an observability platform that is open source, supports metrics, logs, and traces, and can be self-hosted to avoid per-node licensing costs. Which tool stack best fits these requirements?

   • AElasticsearch for logs, with no dedicated metrics or tracing backend.
   • BDatadog for metrics, logs, and traces.
   • CPrometheus for metrics, Loki for logs, and Tempo for traces, all integrated with Grafana.
   • DNew Relic for metrics and traces, with CloudWatch for logs.
   Reveal Solution Discussion Select an answer first
4. 33application · medium

   A traditional monitoring team is transitioning to a cloud native environment with dynamic, short-lived containers. They currently rely on predefined dashboards and threshold-based alerts for known failure modes. They want to adopt an observability approach that better handles unknown failure modes. What is the most important change to make?

   • ACreate more dashboards and alerts for every possible failure mode the team can think of.
   • BReplace all human analysis with automated machine learning that predicts failures before they occur.
   • CCentralize all logs in a single SIEM tool and rely on correlation rules to detect anomalies.
   • DInstrument services to emit rich telemetry (metrics, logs, and traces) and enable ad-hoc exploration of that data to answer new questions.
   Reveal Solution Discussion Select an answer first
5. 34expert · hard

   A platform team has comprehensive monitoring with dashboards and alerts for all known failure modes. After a major incident, they realize they could not diagnose the issue because it was a novel failure mode not covered by any dashboard. Which change would most improve their ability to handle future unknown failures?

   • ACreate more dashboards for every possible metric and log pattern.
   • BHire more monitoring staff to watch dashboards 24/7.
   • CIncrease alerting sensitivity to catch anomalies earlier.
   • DAdopt an observability platform that stores high-cardinality telemetry and supports ad-hoc querying across metrics, logs, and traces.
   Reveal Solution Discussion Select an answer first
6. 35application · medium

   A company is standardizing on OpenTelemetry. They have a legacy service that cannot be modified to add an SDK. Which OpenTelemetry component can collect telemetry from this service without code changes?

   • AThe OpenTelemetry Collector with a receiver for the legacy service's existing protocol.
   • BThe OpenTelemetry SDK for the language the service is written in.
   • CThe OpenTelemetry Operator for Kubernetes.
   • DThe OpenTelemetry CLI tool.
   Reveal Solution Discussion Select an answer first