Automatic Identification of Abnormality in Infrastructure Services

Abstract

In today's world, digital transformation is continuously growing at the exponential rate. Due to this, the infrastructure continues to expand, and the complexity of operations continues to increase. So, monitoring and computing the IT infrastructure data is a very challenging task for the company. Here, IT infrastructure comprises of all the assets required to transfer and assistance with IT services: networks, servers, storage, computer hardware and software, data centers and many more. And to make better monitoring, the evolution of technologies will be around storage, processing big data or even machine learning which has made easier to improve the monitoring in various areas(that is for example in memory usage metric, CPU usage or network connectivity issue which might leads to system fail). IT infrastructure monitoring relies heavily on the logs. Logs can structure or unstructured format, and by using them correctly it can give us the detection of anomaly or an incident or even it can help us to give the future prediction. So, in this report the project agenda is like utilising data science solutions for Infra-monitoring which will study to recognise and categorise abnormal system activity, and by that it will reduce an period it takes to resolve an incident.