Abstract: Systems, methods, and computer program products for identifying a data pattern change anomaly uses a distributing computing environment that processes thousands of different data flows are provided. Numerous data flows are collected from the application computing environment over a configurable time period. The flows are aggregated into aggregated data according to at least one attribute from the flows and without losing information included in the flows. Historical data that includes aggregated data from multiple flows that occurred prior to a time during which the numerous data flows were collected is provided from a distributed disk storage. An anomaly that indicates change in data patterns in the flows is identified by comparing the aggregated data to the historical data using one or more models that are tailored to the numerous flows. An alert that includes an anomaly and a reason for an anomaly is transmitted and recorded in the system.

Abstract: Systems, methods, and computer program products for identifying a data pattern change anomaly uses a distributing computing environment that processes thousands of different data flows are provided. Numerous data flows are collected from the application computing environment over a configurable time period. The flows are aggregated into aggregated data according to at least one attribute from the flows and without losing information included in the flows. Historical data that includes aggregated data from multiple flows that occurred prior to a time during which the numerous data flows were collected is provided from a distributed disk storage. An anomaly that indicates change in data patterns in the flows is identified by comparing the aggregated data to the historical data using one or more models that are tailored to the numerous flows. An alert that includes an anomaly and a reason for an anomaly is transmitted and recorded in the system.