> For the complete documentation index, see [llms.txt](https://atomoh.gitbook.io/kubernetes/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://atomoh.gitbook.io/kubernetes/en/quiz-collection/service-mesh/cilium-service-mesh/observability.md).
# Observability Quiz
This quiz tests your understanding of Hubble, metrics collection, service maps, Golden Signals monitoring, and OpenTelemetry integration.
## Quiz Questions
### 1. Which is NOT a main component of Hubble?
A. Hubble Observer B. Hubble Relay C. Hubble Router D. Hubble UI
Show Answer
**Answer: C. Hubble Router**
**Explanation:** The main components of Hubble are Hubble Observer (embedded in Cilium Agent), Hubble Relay (cluster-wide flow aggregation), Hubble UI (visualization dashboard), and Hubble CLI (command-line interface). Hubble Router is not an existing component.
### 2. What command filters and observes only HTTP traffic in Hubble CLI?
A. hubble observe --type http B. hubble observe --protocol http C. hubble observe --filter http D. hubble observe --layer http
Show Answer
**Answer: B. hubble observe --protocol http**
**Explanation:** The `hubble observe --protocol http` command filters only HTTP protocol traffic. Other protocols (tcp, dns, etc.) can be filtered in the same way.
### 3. What setting needs to be enabled in values.yaml to collect Hubble metrics in Prometheus?
A. hubble.prometheus.enabled: true B. hubble.metrics.enabled C. hubble.export.prometheus: true D. prometheus.hubble: true
Show Answer
**Answer: B. hubble.metrics.enabled**
**Explanation:** To enable Hubble metrics, specify the metric types to collect (dns, drop, tcp, flow, http, etc.) as a list under hubble.metrics.enabled. Also, setting serviceMonitor.enabled: true allows Prometheus Operator to automatically scrape.
### 4. Which is NOT one of the four Golden Signals for monitoring?
A. Latency B. Traffic C. Availability D. Saturation
Show Answer
**Answer: C. Availability**
**Explanation:** The four Golden Signals defined by Google SRE are Latency, Traffic, Errors, and Saturation. Availability is not included in Golden Signals and is measured indirectly through the Errors metric.
### 5. What is the correct command to observe traffic denied by policies in Hubble?
A. hubble observe --denied B. hubble observe --verdict DROPPED C. hubble observe --blocked D. hubble observe --policy-denied
Show Answer
**Answer: B. hubble observe --verdict DROPPED**
**Explanation:** The `--verdict DROPPED` option filters traffic denied by network policies. Conversely, `--verdict FORWARDED` shows allowed traffic.
### 6. Which function is used in PromQL queries to measure HTTP P99 latency?
A. avg() B. histogram\_quantile() C. rate() D. sum()
Show Answer
**Answer: B. histogram\_quantile()**
**Explanation:** Percentile metrics like P99 latency use the histogram\_quantile() function. Example: `histogram_quantile(0.99, rate(hubble_http_request_duration_seconds_bucket[5m]))`. Here, 0.99 means the 99th percentile.
### 7. Which is NOT a main feature provided by Hubble UI?
A. Service Map B. Flow Timeline C. Auto Scaling D. Namespace Filter
Show Answer
**Answer: C. Auto Scaling**
**Explanation:** Hubble UI provides service maps, flow timeline, namespace filter, verdict filter, and L7 details for individual flows. Auto scaling is a workload management feature, not an observability feature.
### 8. What is the correct form of a PromQL query to calculate HTTP error rate in Cilium?
A. hubble\_http\_errors\_total / hubble\_http\_requests\_total B. sum(rate(hubble\_http\_responses\_total{status=\~"5.."}\[5m])) / sum(rate(hubble\_http\_responses\_total\[5m])) C. count(hubble\_http\_errors) / count(hubble\_http\_requests) D. hubble\_http\_error\_rate
Show Answer
**Answer: B. sum(rate(hubble\_http\_responses\_total{status=\~"5.."}\[5m])) / sum(rate(hubble\_http\_responses\_total\[5m]))**
**Explanation:** HTTP error rate is calculated by dividing the count of 5xx responses by the total response count. The rate() function calculates per-second rates, and the status label filter (status=\~"5..") selects only server errors.
### 9. What option is used to observe only traffic destined for a specific service in Hubble?
A. --destination-service B. --to-service C. --target-service D. --svc
Show Answer
**Answer: B. --to-service**
**Explanation:** The `hubble observe --to-service ` command filters traffic destined for a specific service. Conversely, `--from-service` filters traffic originating from a specific service.
### 10. What metric monitors connection tracking table utilization in Cilium?
A. cilium\_ct\_usage B. cilium\_datapath\_conntrack\_active C. cilium\_connections\_total D. cilium\_ct\_table\_size
Show Answer
**Answer: B. cilium\_datapath\_conntrack\_active**
**Explanation:** The cilium\_datapath\_conntrack\_active metric represents the current number of active connections. It can be used with cilium\_datapath\_conntrack\_max to calculate connection tracking table utilization.
### 11. What option is used to receive Hubble output in JSON format?
A. --format json B. -o json C. --json D. --output-type json
Show Answer
**Answer: B. -o json**
**Explanation:** The `hubble observe -o json` command outputs in JSON format. This is useful for piping to tools like jq for additional processing.
### 12. What protocol is used when integrating OpenTelemetry Collector with Hubble?
A. HTTP REST API B. OTLP (OpenTelemetry Protocol) C. Prometheus Remote Write D. StatsD
Show Answer
**Answer: B. OTLP (OpenTelemetry Protocol)**
**Explanation:** Hubble can export flow data to OpenTelemetry Collector using OpenTelemetry Protocol (OTLP). This allows routing data to various backends like Jaeger, Prometheus, Loki, etc.