当前位置：网站首页>Talk about the measurement of "post release problems"

Talk about the measurement of "post release problems"

2022-07-28 17:35:00 【itest_ two thousand and sixteen】

In order to accurately evaluate the product in the market 、 Landing effect on the user side , Enterprises often establish a relatively perfect problem collection and processing mechanism for products . Post release issues （ Sometimes it is also called online problem ） It is an important measurement subject of software product quality . Through the collection of post release problems 、 analysis , On the one hand, we can have a clearer understanding of the quality performance after product release , On the other hand, it can also promote the collaboration of front and rear organizations and the optimization of workflow .

Generally speaking , We will look at the quality performance of a product after release from the number of problems after release , But this is not objective and it is impossible to compare the quality performance of several products horizontally . for instance A The products were delivered to ten customers 1 In total, there are 20 Post release questions ;B The product was delivered to 100 Customers ,1 In total, there are 50 Post release questions . Obviously , We can't use the index of the number of problems after release A The quality of our products is better than B product , Because in the case of unit customers B The number of problems after the release of the product is significantly less than A product . But this can explain B Product quality is better than A Products ？ Not necessarily. . such as ：A The complexity of product development may be higher , When the after-sales service system is consistent, the quality problem is proportional to the complexity ;B The guarantee measures during product delivery are more complete 、 The service is of a high standard, which greatly reduces the probability of problems being exposed .

therefore , We introduced “ Post release problem density ” This indicator . The calculation method of this index is ： Post release problem density = Number of issues after release / Product size . This formula is not difficult to understand , The difficulty is to figure out what is the product scale , And how to calculate the product scale . Before discussing how to define and account for product size , We need to understand a concept , That is, no matter how you define and calculate the product scale, you can't accurately calculate the problem density after release , Moreover, the improvement of accuracy will also lead to a significant increase in the overall measurement cost . Reduce post release problems , It is the guarantee of various unit output , The same product , The larger the scale of various units , The lower the probability of problems （ Be careful , This is a probability event , Don't take extreme cases ）. We develop our products from design to R & D , From research and development to release 、 The input of each link delivered to the customer is added together as the product scale , It's not just human input , Also contains “ matter ” Investment . For example, some customers purchase hard loads for a more robust system , A large number of network security equipment have been purchased for system security . According to the above formula , We can get the number of problems per unit input after release as the problem density . Is this algorithm reliable ？ Obviously not , Because accounting is too complicated .“ Post release problem density ” The indicators look good , But the operability is very low .

A friend will ask , I only compare the quality of a product this month compared with that after its release last month , Is there no need for “ Post release problem density ” This indicator is , Just look at the quantity directly ？ I can only say that unless your sales volume has been unchanged and your product has been in the basic operation and maintenance period .

Return to nature , The collection and analysis of post release problems are not for quality comparison between products , But to promote product quality optimization and improve workflow . So which indicators of the post release question are more valuable ？ For example, the problem response time 、 Processing time 、 Verification duration 、 Life cycle length . Beyond the indicators , Analysis of post release problems 、 replay 、 It is more important to solidify the solution into the workflow than the measurement itself .

We have several definitions about the problem life cycle for reference ：

Problem life cycle ： The whole cycle of the problem from raising to closing . Used to measure production and research - The overall efficiency of the delivery team in dealing with problems . If the life cycle of a large number of problems is too long , It shows that there may be efficiency problems in the cooperation from delivery to production and research .

Problem response time ： The length of time from problem recording to problem solving .

Problem solving time ： The time from the beginning of problem solving to submission for verification . It is used to measure the problem handling efficiency of the production and research team , If the problem is solved for a long time , It indicates that there may be problems such as efficiency problems or insufficient manpower .

Problem verification duration ： Problems range from submitting validation to closing . Used to measure the efficiency of problem verification . If the problem is important , achieve P1 or P2, But verification takes 2、3 God , Then it indicates that there may be problems, and the grading is random 、 The verifier is not familiar with the system .