8th CMG Bangalore Event – Quick Quiz for Performance Testers

During our 8th CMG Bangalore Event, we had planned for this quick quiz for Performance Testers (as a filler activity between the technical sessions).

Feel free to take up the quiz & share your answers with us at the mail id provided below.

Entries with all right answers will be eligible for goodies during our next CMG Bangalore regional event 🙂

Question #1

  • Assume you are running a 4000 users test , where the below measurements are made during steady state period. Throughput is about 120 transactions/second, total response time of all 10 transactions in the script is about 20 seconds (each transaction’s average response time is about 2 seconds approximately) & configured total think time is about 25 seconds. Is this a valid test?
    • Option A : Valid Test
    • Option B : InValid Test
    • Option C : Valid , if the server utilization levels are healthy
    • Option D : Not enough details to confirm the test validity

Question #2

  • You are benchmarking a web system for its target load & certifying the system that it will support 300 TPS load with CPU service demand of 0.001. Lets assume the server CPU threshold is 80%. Few weeks after project completion, you get a call from client , “am expecting 1000 TPS load , can the server handle this load. Am just looking for a quick ball park estimate as we don’t have time to run the test” ? Which of the below is your style ?
    • Option A : Double the current CPU capacity
    • Option B : Current CPU capacity should be good enough
    • Option C : 1.25 times the current CPU capacity is required + lets add additional buffers
    • Option D : 1.5 times the current CPU capacity is required + lets add additional buffers

Question #3

  • Which of the below is a Application Performance Management (APM) Tool ?
    • Option A : Retrace
    • Option B : Scout
    • Option C : SteelCentral
    • Option D : Nastel
    • Option E : None of the above
    • Option F : All of the above

Question #4

  • What is BlockChain ?
    • Option A : Network Terminology
    • Option B : Distributed database
    • Option C : Programming Language
    • Option D : IOT Terminology
    • Option E : None of the above

Question #5

  • Which of the below refers to OLTP workload benchmark in TPC that focuses on database?
    • Option A : TPC-DS
    • Option B : TPC-V
    • Option C : TPC-E
    • Option D : TPC-H
    • Option E : None of the above

Question #6

  • Which of the following is the most popular protocol in IOT environments ?
    • Option A : AMQP
    • Option B : CoAP
    • Option C : MQTT
    • Option D : XMPP
    • Option E : None of the above

Question #7

  • A system needs to be load tested for a 1000 users & the target throughput SLA is about 100 TPS & response time SLA for its 3 transactions is 2 seconds. What is the best think time to be configured in the script to run a realistic load test to achieve this SLA?
    • Option A : 0 seconds
    • Option B : ~6 seconds
    • Option C : ~4 seconds
    • Option D : It depends
    • Option E : None of the above

Question #8

  • During NFR Gathering Phase, you understand that, the E-commerce business expects a load of 1000 users doing 4  transactions in a use case , with response time SLA of 1 second each, should yield 500 transactions per second (to meet their business targets) . On an average, the total think time to perform the business use case with 4 transactions is about 6 seconds. Business insists to do a load test & confirm whether the system can meet this NFR ?
    • Option A : You will plan to run the test to confirm
    • Option B : You bet this is not a Valid NFR
    • Option C : You say this is a Valid NFR
    • Option D : Skip the meeting

Question #9

  • While load testing a 3 tier web application, the system is able to achieve 200 TPS throughput & you have calculated the CPU service demand for web, application & DB servers as 0.001, 0.003 & 0.002 respectively. Lets assume CPU threshold is 85%. With this details, is it possible to predict the maximum achievable system throughput without conducting further tests ?
    • Option A : No, additional load test has to be run
    • Option B : System can achieve ~280 transactions
    • Option C : System can achieve ~425 transactions
    • Option D : System can achieve ~850 transactions
    • Option E : Incomplete inputs.

Question #10

  • A dedicated back-end batch processing system is seen to have CPU utilization of about 96% & the average turnaround time for the scheduled job is 7 minutes. Is this healthy system?
    • Option A : There is a scope for reducing the response time
    • Option B : System is not healthy as CPU utilization of 96% is too high
    • Option C : Incomplete inputs
    • Option D : Yes, System can be considered healthy as it’s a dedicated batch processing system.

Mail your answers / queries to ramya.moorthy@elitesouls.in