Performance testing: The process of using automated tools to simulate different scenarios to test and evaluate various performance indicators of the software. Includes load and stress testing
Load test: The performance of the system under different loads. Through this test, the performance change curve of the system under different loads can be found, so as to find the inflection point of performance. For example, during load testing, when only the number of concurrent users is continuously increasing, observe the changing patterns of various performance indicators, find the maximum TPS that the system can achieve, and observe the average response time of system processing at this time, and the performance of various system resources and hardware resources. Consumption status.
Stress test: The performance of the system under high load. This test is mainly to find the maximum load that the system can withstand and the throughput rate of the system at this time. Through this test, it can also be found whether the system will collapse and fail under ultra-high load. access, and whether system performance can automatically recover after the system load is reduced.
Benchmark test: A test performed on the version under development of the system under test, collecting various performance indicators as a comparison of the performance of later versions.
Stability test: Test the system for a long time with normal load or slightly higher than normal load to detect whether the system can run stably for a long time and whether the various performance indicators of the system will change significantly over time.
Scalability test: Usually used for newly launched systems or newly built system environments. By first testing the processing capabilities of a single server, and then slowly increasing the number of servers, it tests whether there is any loss of processing capabilities of a single server in a cluster environment. Whether the performance processing capability of the cluster environment can show a steady increase.
1 Performance test scenario
Single scenario: refers to a scenario involving only a single business process. The purpose is to test whether the system's single business processing capability meets expectations, and to obtain performance indicators such as the maximum TPS and average response time under normal system resource utilization.
Business scenario: refers to the business processing process of the system, describing specific user behavior, and analyzing user behavior to draw out different business scenarios. It is an important source of test scenario design during performance testing.
Mixed scenario: Multiple business processes are designed in the test scenario, and the proportion of each business process in the mixed business process will be different. The proportion is generally set according to the actual business process and meets the actual business needs as much as possible. The purpose of this test scenario is to test whether the system's mixed service processing capabilities meet the expected requirements, and to evaluate the maximum mixed service processing capacity of the system.
2 Performance test indicators
To measure the performance of a system, some performance indicators are used for analysis and description in performance testing. The following are some of the most commonly used performance indicators.
Transaction: It is a customized operation or a collection of operations. For example: a login page, enter the user name and password, jump from clicking the login button to the login completion page, and the new page is completely loaded, this operation is defined as a transaction.
PV: Page View refers to the page views or clicks. Every time a user visits any page in the system or website, it will be recorded once. If the user visits the same page multiple times, the number of visits will be accumulated +1. PV is generally an important indicator to measure the performance capacity of e-commerce websites.
UV: Unique Visitor is an independent visitor of the system. A computer client (IP) that accesses the system website is a visitor.
Click-through rate (hit): The number of page clicks per second. The number of requests submitted by the client to the server per second. In performance testing, static requests (such as Js, css, image files, etc.) are generally not tested. Static requests are cached and do not require server processing. So hit usually refers to dynamic request.
Buried points: Also known as event tracking, that is, analyzing user behavior. Which page did the user click on and how many times did he click? How long did you stay on which page? What item are you searching for? All will be counted. Then do big data analysis to analyze your age level and what you want to buy recently based on your search, so as to accurately push what you need and increase the transaction rate. (The buried data is usually done by operations for big data analysis)
Number of concurrent users: In real user operations, there will be a certain interval between each adjacent operation of the user (called user thinking time), so concurrent users are divided into absolute concurrency and relative concurrency. In terms of performance indicators alone, the number of concurrent users of the system refers to the total number of users that the system can carry at the same time and use system functions normally.
------ Absolute concurrency: refers to the number of concurrent users who make requests to the server at the same time at a certain point in time. --- Flash sale items
------ Relative concurrency : refers to the total number of concurrent users making requests to the server within a period of time.
Response time RT: response time of transaction processing. The difference between the time a request or an operation is issued and the time it receives a response from the server
TPS: Transactions processed per second. The number of client requests the system handles per unit time
QPS: Query rate per second. A measure of how much traffic a specific query server handles within a specified period of time, regardless of external factors.
TPS: Transactions processed per second. The number of client requests the system handles per unit time
Throughput: refers to the number of client requests processed by the system per unit time. From different perspectives, throughput can be calculated in different ways.
(1) From a business perspective: Throughput can be measured and calculated by the number of requests/s, the number of pages/s, etc.
(2) From a network perspective: Throughput can be measured and calculated in bytes/s.
(3) From an application perspective: The throughput indicator reflects the pressure on the server and the load capacity of the system.
The throughput of a system is generally closely related to the consumption of CPU, bandwidth, IO and memory resources by a request processing.
Resource overhead: The consumption of system resources by each request or transaction, used to measure the degree of resource consumption by requests or transactions.
Idempotence:
That is, the results of one request or multiple requests initiated by the user for the same operation are consistent, and there will be no side effects caused by multiple clicks. For example, after a user purchases XX product, the payment is deducted successfully for the first time, but the network is abnormal when the result is returned, and the money has been deducted at this time. When the user pays again, the second deduction will be made and the result returned is successful. It is equivalent to the user paying twice when purchasing XX goods.... This situation requires idempotence testing. If the first deduction has been made, the business needs to be rolled back and no longer be deducted.
Finally: The complete software testing video tutorial below has been compiled and uploaded. Friends who need it can get it by themselves [guaranteed 100% free]
Software Testing Interview Document
We must study to find a high-paying job. The following interview questions are the latest interview materials from first-tier Internet companies such as Alibaba, Tencent, Byte, etc., and some Byte bosses have given authoritative answers. After finishing this set I believe everyone can find a satisfactory job based on the interview information.
