How to deal with occasional bugs?

In the course of daily testing, it is inevitable to encounter some strange problems , How to treat this kind of BUG？

My approach ：

（1） Try to reproduce it many times first , Find the path to the problem

result 1： Successfully found the required path ,game over

result 2： No mandatory path found , turn （2） Handle

Try to find the necessary path ：

First step ： Recall the operation steps of the problem , Try to repeat the previous steps

The second step ： If you press the first step, you will still not be able to reproduce , Start the screen recording software and repeat 【 First step 】 Operate multiple times , Once it happens again, observe it again by playing back the recorded steps , Find out the possible operation scenarios of the problem

The third step ：【 For more serious problems 】 When the manual operation cannot be reproduced , With the help of some automated tools （ Such as writing automation script or key wizard ）, Long time automated script execution 【 First step 】 Operate for a period of time to reproduce

（2） Record and track such problems , As usual BUG Submit to defect tracking Library

BUG Submit notes ： The steps should be as detailed as possible , Indicate the 【 Epiphany 】 problem , The frequency of occurrence 、 Log information when problems occur 、 If there is an occasional screen recording or screenshot, it is in BUG Attached

（3） During the follow-up test , Pay attention to whether such problems recur

Once it happens again （ Especially the serious problem ） after , Keep the environmental site , Pull the corresponding developer to immediately troubleshoot the problem .