Intelligency log analyses on in-memory database systems

Data of all time and ages have always been valuable; it has pioneered the development of the progress of science and society. In the information and technology age we live in, the value of data is gaining more importance every day. Today, one of the most fundamental issues that make the extremely important data, the data can be processed quickly and has to be better analyzed. More meaningful data on analysis results produced and thus, the data can be used much more efficiently. In this study, data of logs that occur in the operation process of the system and application in the In-Memory Database Systems keeping ANN algorithms SOM (Self Organizing learning) analyzing the data held using the algorithm and approach to the detection of anomalies with the system is based. Anomaly detection for the analysis of data, supervised learning (supervised) ANN algorithms will not be sufficient alone, unsupervised learning (unsupervised) neural network has declared its opinion that the requirements of the use. For rapid analysis of large volumes of data on the other hand, In-Memory has been given for the use of database systems. Developed a prototype application is described in detail on the subject. In practice, In-Memory database held in the table belonging to a Windows server in the IT system and application log data received via WMI is used. First SOM data were analyzed using the algorithm, then it made based on the anomaly detection output data and abnormality is rated levels.