> 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/linkerd/traffic-management.md).
# Traffic Management Quiz
This quiz tests your understanding of Linkerd traffic management.
## Quiz Questions
### 1. What cannot be configured per-route in a ServiceProfile?
A. Timeout B. Retryability C. Load balancer algorithm D. Path condition
Show Answer
**Answer: C. Load balancer algorithm**
**Explanation:** ServiceProfile can configure timeout, retryability (isRetryable), and path conditions (method, pathRegex) per route. The load balancer algorithm is a Linkerd global setting, using EWMA.
### 2. What load balancing algorithm does Linkerd use?
A. Round Robin B. Least Connections C. EWMA (Exponentially Weighted Moving Average) D. Random
Show Answer
**Answer: C. EWMA (Exponentially Weighted Moving Average)**
**Explanation:** Linkerd uses the EWMA algorithm to prefer endpoints with faster response latency. It adapts to endpoint state in real-time and automatically reduces traffic to slow endpoints.
### 3. What standard specification does TrafficSplit follow?
A. CNCF B. SMI (Service Mesh Interface) C. OpenAPI D. gRPC
Show Answer
**Answer: B. SMI (Service Mesh Interface)**
**Explanation:** TrafficSplit is a CRD that follows the SMI (Service Mesh Interface) standard. SMI defines common interfaces for service meshes to provide compatibility between different mesh implementations.
### 4. What does a retryBudget retryRatio of 0.2 mean?
A. Only 20% of all requests are retried B. Only 20% of failed requests are retried C. Up to 20% additional retries allowed relative to original requests D. Retry budget resets every 20 seconds
Show Answer
**Answer: C. Up to 20% additional retries allowed relative to original requests**
**Explanation:** A retryRatio of 0.2 allows up to 20% additional retries relative to the original number of requests. Example: Up to 20 additional retries allowed for 100 requests. This prevents overload from retries.
### 5. Which is NOT a method to auto-generate a ServiceProfile?
A. Generate from OpenAPI/Swagger spec B. Generate from live traffic tap C. Generate from Protobuf definition D. Auto-generate from Kubernetes Service
Show Answer
**Answer: D. Auto-generate from Kubernetes Service**
**Explanation:** ServiceProfiles can be generated using `linkerd profile --open-api`, `linkerd viz profile --tap`, and `linkerd profile --proto` commands. They are not auto-generated from Kubernetes Services and must be explicitly defined.
### 6. What should the sum of TrafficSplit backend weights be for canary deployment?
A. Must be exactly 100 B. Must be exactly 1 C. Any value works (calculated as ratio) D. Must be exactly 1000
Show Answer
**Answer: C. Any value works (calculated as ratio)**
**Explanation:** TrafficSplit weights are calculated as relative ratios. weight: 90 and weight: 10 is equivalent to weight: 9 and weight: 1. The sum doesn't need to be 100.
### 7. What routing condition is NOT supported in HTTPRoute (Gateway API)?
A. Header-based routing B. Path-based routing C. Cookie-based routing D. Source IP-based routing
Show Answer
**Answer: D. Source IP-based routing**
**Explanation:** HTTPRoute supports header, path, method, and cookie (via headers) based routing. Source IP-based routing is outside the scope of L7 routing and is handled by NetworkPolicy or other mechanisms.
### 8. What metrics server is used when integrating Flagger with Linkerd?
A. Metrics Server B. Prometheus C. InfluxDB D. Datadog
Show Answer
**Answer: B. Prometheus**
**Explanation:** Flagger retrieves metrics (success rate, latency, etc.) from Linkerd Viz's Prometheus for canary analysis. When installing Flagger, connect with `--set metricsServer=http://prometheus.linkerd-viz:9090`.
### 9. What happens on a route where ServiceProfile isRetryable is false?
A. All requests fail B. No retries occur C. Timeouts are ignored D. Route is disabled
Show Answer
**Answer: B. No retries occur**
**Explanation:** isRetryable: false means requests on that route will not be retried even if they fail. This is suitable for non-idempotent operations like POST requests. The request itself is processed normally.
### 10. How is the Circuit Breaker pattern implemented in Linkerd?
A. Circuit Breaker CRD B. Failure Accrual C. Rate Limiter D. Timeout Policy
Show Answer
**Answer: B. Failure Accrual**
**Explanation:** Linkerd implements the circuit breaker pattern through failure accrual. On consecutive failures, it temporarily excludes the endpoint, retries with exponential backoff, and returns to normal state on success.
### 11. How do you send traffic to a mirror service without traffic splitting?
A. Use TrafficMirror CRD B. Call mirror service DNS directly C. All traffic is automatically mirrored D. Linkerd doesn't support traffic mirroring
Show Answer
**Answer: B. Call mirror service DNS directly**
**Explanation:** Linkerd itself doesn't have traffic mirroring functionality like Istio. Multi-cluster mirror services (e.g., web-west) must be called directly via DNS or configured with TrafficSplit weights.
### 12. What happens on a route where ServiceProfile timeout is not set?
A. Default 5-second timeout applies B. No timeout (unlimited) C. Request fails immediately D. Global timeout applies
Show Answer
**Answer: B. No timeout (unlimited)**
**Explanation:** Routes without a timeout specified in ServiceProfile wait indefinitely without timeout. This is suitable for streaming or long-running operations, but explicitly setting timeouts is generally recommended.