In software testing, the analysis of logs is designed to check and monitor the application work. Also logs help to fix some errors. Logs are the text lines that include time and name of particular operations that occur in different environments - application itself or the system that runs an app. Logs include the messages of several types: informational, warning and error.
Logs help to specify who uses an application and how often it is used. Moreover, testers can define whether the session was successful, and to detect possible mismatches or mistakes a users of the app can face.
Log analysis: qualitative approach
The log size shows if the app is used lightly or heavily. If all logs are stored on one hard drive, the number of users is smaller. The larger logs are, the bigger number of users an application has. But mind that it depends on the log mode. On the basis of log size, the testers define the age of a given application. The application with the logs of several entries can not be released ten years ago. And vice versa, if the app has a million of logs that it was not installed an hour ago. But do not forget about Debug mode. Depending on, the speed of log generation identifies the number of active users. During web testing, the specialists analyze the log generation patterns if any. The matter is that some apps are actively used only during working hours while others are not.
Log analysis: quantitative approach
The quantitative data assist testers in evaluating the quality of product functionality. For example, four thousand errors logged in the past week mean poor quality of an application. The analysis of logs is a part of functional testing as well as performance testing. But what to do if the number of logs is very big? How to track errors and warnings? Of course, manual checking of logs takes much time and may bring no results. The structure of log itself solves the problem. In general, logs have somehow standardized structure that allow the tester to parse them using special scripts. So it is possible to pull out any required data.
Logs reflects the real state of affairs
The real user data is a must-have for every type of software testing. If checking is based only on assumptions if will not be resultative. Logs demonstrate how the application is used, and which parts of them are utilized most actively. Testers are able to evaluate the application behavior in real environment. The log analysis helps to design new scenarios for further web or mobile testing. And probably, the biggest advantage of logs for both developers and tester is that they prove that people really use the product :)