Log files contain enormous numbers of events projected by several system(s) activities. Each event is normally represented by one line in the log file. A log file is an excellent technique to maintain information about the run time of software.  Hence, it is convenient to use the log file to analyze the behavior of a system or detect anomalies. Manual analysis is impractical due to the massive log file size besides the necessity for subject matter knowledge. In this project, we will use machine-learning techniques to overcome this issue