Performance Metrics interview questions — Easy
250 easy-level Performance Metrics questions from our Performance Testing bank, each with the correct answer and an explanation of why it is correct.
What this covers
Performance Metrics appears throughout Performance Testing interviews. At easy level, interviewers are typically checking that you know the fundamentals and can describe them precisely — definitions, defaults, and the basic mechanics. Work through these, then explain your answer out loud; the second part is what interviews actually test.
20 example questions
1. In Metrics, what does connection pool exhaustion do or refer to?
- A. Measurement error from pausing load when responses stall.
- B. Transactions processed per second.
- C. JVM stop-the-world event inflating tail latency.
- D. Bottleneck when no free connections remain.correct
Why: connection pool exhaustion: Bottleneck when no free connections remain.
2. In the context of Metrics, time to first byte (TTFB) is best described as:
- A. Transactions processed per second.
- B. Service-level objective, e.g., 99% of requests under 300ms.
- C. JVM stop-the-world event inflating tail latency.
- D. Delay until the first response byte arrives.correct
Why: time to first byte (TTFB): Delay until the first response byte arrives.
3. Which of the following best describes p50 (median) in Metrics?
- A. Value splitting the faster and slower half of requests.correct
- B. Bottleneck when no free connections remain.
- C. Measurement error from pausing load when responses stall.
- D. Service-level objective, e.g., 99% of requests under 300ms.
Why: p50 (median): Value splitting the faster and slower half of requests.
4. Which of the following best describes connection pool exhaustion in Metrics?
- A. Transactions processed per second.
- B. Bottleneck when no free connections remain.correct
- C. L = λ × W relating concurrency, arrival rate, and time in system.
- D. Reference run against which future results are compared.
Why: connection pool exhaustion: Bottleneck when no free connections remain.
5. In Metrics, what does time to first byte (TTFB) do or refer to?
- A. Delay until the first response byte arrives.correct
- B. Measurement error from pausing load when responses stall.
- C. Reference run against which future results are compared.
- D. Bottleneck when no free connections remain.
Why: time to first byte (TTFB): Delay until the first response byte arrives.
6. What is the purpose of connection pool exhaustion in Metrics?
- A. Service-level objective, e.g., 99% of requests under 300ms.
- B. Bottleneck when no free connections remain.correct
- C. Value splitting the faster and slower half of requests.
- D. Maximum sustainable throughput within response-time targets.
Why: connection pool exhaustion: Bottleneck when no free connections remain.
7. In Metrics, what does garbage collection pause do or refer to?
- A. Value splitting the faster and slower half of requests.
- B. Measurement error from pausing load when responses stall.
- C. Maximum sustainable throughput within response-time targets.
- D. JVM stop-the-world event inflating tail latency.correct
Why: garbage collection pause: JVM stop-the-world event inflating tail latency.
8. Which of the following best describes time to first byte (TTFB) in Metrics?
- A. Transactions processed per second.
- B. Measurement error from pausing load when responses stall.
- C. L = λ × W relating concurrency, arrival rate, and time in system.
- D. Delay until the first response byte arrives.correct
Why: time to first byte (TTFB): Delay until the first response byte arrives.
9. Which of the following best describes Little's Law in Metrics?
- A. Maximum sustainable throughput within response-time targets.
- B. JVM stop-the-world event inflating tail latency.
- C. L = λ × W relating concurrency, arrival rate, and time in system.correct
- D. Service-level objective, e.g., 99% of requests under 300ms.
Why: Little's Law: L = λ × W relating concurrency, arrival rate, and time in system.
10. Which of the following best describes baseline test in Metrics?
- A. JVM stop-the-world event inflating tail latency.
- B. Reference run against which future results are compared.correct
- C. Delay until the first response byte arrives.
- D. L = λ × W relating concurrency, arrival rate, and time in system.
Why: baseline test: Reference run against which future results are compared.
11. Which of the following best describes coordinated omission in Metrics?
- A. Measurement error from pausing load when responses stall.correct
- B. L = λ × W relating concurrency, arrival rate, and time in system.
- C. Reference run against which future results are compared.
- D. Maximum sustainable throughput within response-time targets.
Why: coordinated omission: Measurement error from pausing load when responses stall.
12. In Metrics, what does SLO do or refer to?
- A. JVM stop-the-world event inflating tail latency.
- B. L = λ × W relating concurrency, arrival rate, and time in system.
- C. Service-level objective, e.g., 99% of requests under 300ms.correct
- D. Maximum sustainable throughput within response-time targets.
Why: SLO: Service-level objective, e.g., 99% of requests under 300ms.
13. Which of the following best describes capacity in Metrics?
- A. Measurement error from pausing load when responses stall.
- B. Maximum sustainable throughput within response-time targets.correct
- C. Service-level objective, e.g., 99% of requests under 300ms.
- D. JVM stop-the-world event inflating tail latency.
Why: capacity: Maximum sustainable throughput within response-time targets.
14. What is the purpose of coordinated omission in Metrics?
- A. JVM stop-the-world event inflating tail latency.
- B. Reference run against which future results are compared.
- C. Measurement error from pausing load when responses stall.correct
- D. L = λ × W relating concurrency, arrival rate, and time in system.
Why: coordinated omission: Measurement error from pausing load when responses stall.
15. In Metrics, what does p50 (median) do or refer to?
- A. L = λ × W relating concurrency, arrival rate, and time in system.
- B. Bottleneck when no free connections remain.
- C. Value splitting the faster and slower half of requests.correct
- D. Service-level objective, e.g., 99% of requests under 300ms.
Why: p50 (median): Value splitting the faster and slower half of requests.
16. What is the purpose of time to first byte (TTFB) in Metrics?
- A. Delay until the first response byte arrives.correct
- B. Measurement error from pausing load when responses stall.
- C. JVM stop-the-world event inflating tail latency.
- D. Value splitting the faster and slower half of requests.
Why: time to first byte (TTFB): Delay until the first response byte arrives.
17. In Metrics, what does capacity do or refer to?
- A. Bottleneck when no free connections remain.
- B. Maximum sustainable throughput within response-time targets.correct
- C. Value splitting the faster and slower half of requests.
- D. JVM stop-the-world event inflating tail latency.
Why: capacity: Maximum sustainable throughput within response-time targets.
18. What is the purpose of capacity in Metrics?
- A. L = λ × W relating concurrency, arrival rate, and time in system.
- B. Value splitting the faster and slower half of requests.
- C. Maximum sustainable throughput within response-time targets.correct
- D. Bottleneck when no free connections remain.
Why: capacity: Maximum sustainable throughput within response-time targets.
19. What is the purpose of garbage collection pause in Metrics?
- A. Bottleneck when no free connections remain.
- B. Transactions processed per second.
- C. JVM stop-the-world event inflating tail latency.correct
- D. Delay until the first response byte arrives.
Why: garbage collection pause: JVM stop-the-world event inflating tail latency.
20. What is the purpose of SLO in Metrics?
- A. Bottleneck when no free connections remain.
- B. Delay until the first response byte arrives.
- C. Service-level objective, e.g., 99% of requests under 300ms.correct
- D. Transactions processed per second.
Why: SLO: Service-level objective, e.g., 99% of requests under 300ms.