[Performance Test] LoadRunner Results Analysis –tps

[Performance Test] LoadRunner Results Analysis –tps

For the relationship between throughput rate and TPS, how to use this in the analysis of results, after personal experience and friends to discuss, put forward the following suggestions to guide, welcome the colleague to correct.

Related definitions

• response time = Network response time + Application response time
• response time = (N1+N2+N3+N4) + (A1+A2+A3)
• tps:trasaction per second is the number of transactions per second. It is the unit of measurement for software test results. A transaction is the process by which a client sends a request to the server and the server responds. The client starts the time when the request is sent and ends the timer after receiving the server response, which is used to calculate the elapsed and the number of transactions completed, and ultimately the information is used to estimate the score. The client uses the weighted covariance averaging method to calculate the client's score, and the test software uses this information from the client to calculate the overall TPS score on the server side using the weighted covariance averaging method. Generally speaking, the TPS of the system depends on the TPS of the module with the minimum processing power of the system transaction, XP 10-100
• hps:hit per second is the number of clicks/sec, referring to the user's link to the Web page in a second time, Submit button for the sum of clicks. Generally proportional to TPS, is a major indicator of the B/s system
• throughput/s: Throughput, which refers to the number of requests per second that the system processes, and can be understood as the amount of feedback received by the customer within a unit of time
• Throughput rate: The number of bytes returned from the server per second during the test.

from the definition, if the TPS is small, but the throughput ratio is large, the server's return page file (number of bytes ) is relatively large, at this time according to the page breakdown chart, if there is a page problem, consider page compression.  

should also be based on a1-a3, n1-n3 practical consideration.  

If the A1 or A3 is larger, the webserver processing may be a problem, if the A2 is relatively large, then dbserver processing has a problem, recommended SQL optimization.

When the system pressure increases (or increases the number of concurrent users ), the throughput rate and the change curve of TPS are broadly consistent, the system is basically stable.  

If the pressure increases, the throughput curve increases to a certain extent after the change is slow, even flat, while the TPS tends to flat, viewing system resources use, if the resource utilization is low, the server hardware resources are not a problem, view network traffic, the network bandwidth is estimated to be problematic.  

Similarly, if the CTR/tps curve changes slowly or smoothly , click- through rate (number of requests per second ) if the pressure increases, tend to flat, it is likely that the server response time increased, observe the server resource usage, determine whether it is a server problem.  

TPS is the unit of measurement for software test results. A transaction is the process by which a client sends a request to the server and the server responds. The client starts the time when the request is sent and ends the timer after receiving the server response, which is used to calculate the elapsed and the number of transactions completed, and ultimately the information is used to estimate the score. The client uses the weighted covariance averaging method to calculate the client's score, and the test software uses this information from the client to calculate the overall TPS score on the server side using the weighted covariance averaging method .  

In general, the performance of the evaluation system is measured in terms of the number of technical transactions completed per second. The overall processing capacity of the system depends on the TPS value of the lowest processing capacity module . According to experience, the processing capacity of the application system generally requires about 10-100. TPS in different application Systems has a very large difference, and it is generally necessary to make accurate estimation by performance test.

Empirical Analysis:

1, TPS standard deviation/tps average>8%, or <2% system has a performance bottleneck

2,when increasing the pressure of the system (or increasing the number of concurrent users ), the throughput rate and the TPS change curve is proportional to the change, the system is basically stable

3, if the pressure increases, the throughput rate curve increases to a certain degree after the change is slow, even flat, while TPS tends to flat, viewing system resources use, if the resource utilization ratio is high, then the server hardware resources have problems, need to expand hardware or optimize the application. Conversely, the server hardware resources are not a problem, look at network traffic, estimated network bandwidth problems.

4, Ctr /tps curve changes slow or flat , it is likely that the server response time increased, observe the server resource usage, determine whether it is a server problem or application problems

[Performance Test] LoadRunner Results Analysis –tps (RPM)