Oracle, an enterprise solutions company that develops database and middleware software, application software, and hardware systems, is focused on driving transformation in the cloud. As such, it understands that performance, scalability, reliability, and security are the cornerstones for success in the cloud market.
The company is using big data principles and technology, including Oracle Database, Oracle Enterprise Manager, Oracle WebLogic Server and Oracle Big Data Appliance, to deliver new levels of security and performance to the cloud.
First, Oracle is enabling near-real-time analysis of its cloud systems’ behavior by various means, including Oracle WebLogic Server Diagnostic Framework, Oracle Enterprise Manager Monitoring Templates and corrective actions, and by collecting and analyzing hundreds of metrics and log files on a near-real-time basis. This enables Oracle to detect sudden changes (anomalies) in security, performance, scalability, and availability characteristics of the system. Second, Oracle is enabling intermediate-term analysis of the systems’ behavior by collecting thousands of log files into an archive and analyzing them for anomalies in systems’ behavior, relating them back to changes in workload, system outages, and system upgrades, and using logs for defect detection and resolution in the cloud services. Third, log files from the archive are then shipped to Oracle Big Data Appliance for long-term storage, statistical analysis, trend computation, anomaly detection, and predictive analytics.
Oracle Big Data Appliance includes Apache Hadoop big data technology. It can be used to analyze millions of log files to identity potential nefarious activity, such as attempts at unauthorized access. For example, it can rapidly identify low-frequency attack patterns over a long period of time, such as an individual unsuccessfully logging in three or four times a week for a long time, which might indicate fraudulent activity. While multiple failed log-in attempts at the same time set an immediate flag, more subtle and less frequent failed attempts, which might also signal an issue, have largely gone unidentified. Today, Oracle can identify events that warrant further investigation or immediately suspend access from a specific IP address, to improve overall security of Oracle Managed Cloud Services offerings.
Further, Oracle is using big data to improve cloud environment performance and ensure service-level-agreement compliance, which includes latency of less than one second. For example, it can now rapidly and effectively track and analyze how Java Virtual Machines in Oracle Public Cloud are using heap memory. If they spot a trend, Oracle can dig deeper and determine the best path forward before performance is impacted.
In addition, the company is planning to use big data to enable dynamic resource management. It will analyze, identify, and even predict usage trends, such as seasonal, monthly, or weekly spikes, and then proactively spin up additional instances to accommodate heavier loads. This approach will also cut down on the number of last-minute fire drills, enabling the IT team to focus more of its attention on strategic initiatives and improving its overall work experience.