Business analytics enable organizations worldwide to gain better business insights, make better business decisions, and deliver better business results. In fact, business analytics and related technology initiatives consistently rank among the top three investment priorities of IT executives today.
But the vision of delivering fast, interactive, and insightful analytics has remained elusive for many organizations. Most enterprise IT organizations continue to struggle with the delivery of timely and actionable analytics because they lack an integrated technology solution. And they spend a disproportionate amount of time and resources trying to integrate a variety of hardware, software, storage, and networking products in their quest to provide a complete analytics solution. What's more, the same process of putting together an end-to-end integrated solution is repeated again and again for every new analytics project. And yet, this highly inefficient process remains status quo for most businesses. Oracle has the solution.
Oracle Exalytics In-Memory Machine is the world's first end-to-end integrated solution specifically designed to deliver high-performance analysis, modeling, and planning. Built using market-leading business intelligence software, in-memory database technology, and industry-standard hardware including state-of-the-art processors from Intel®, Oracle Exalytics is an engineered system that delivers answers to your most pressing business questions with unmatched speed, intelligence, simplicity, and manageability. With an integrated and optimized hardware and software analytics solution, Oracle Exalytics can empower your IT organization to deliver scalable and repeatable speed-of-thought analytics at a fraction of the time and cost of traditional, build-your-own analytics solutions.
Oracle Exalytics can be deployed in existing IT environments individually or in conjunction with Oracle Exadata and/or Oracle Exalogic systems to enable extreme analytics and a best-in-class user experience.
Oracle Exalytics is an assured in-memory solution that utilizes the enhanced reliability, availability, and serviceability (RAS) capabilities of the Intel® Xeon® processor E7-8800/4800/2800 product families at both the hardware and software level to guarantee operational robustness and efficiency.
With unmatched speed and scalability, Oracle Exalytics delivers extreme performance for existing analytical and enterprise performance management applications and enables a new class of intelligent applications like near real-time yield management, demand forecasting, inventory management, pricing optimization, profitability management, churn analysis, rolling forecast, virtual close, and more.
Companies worldwide are using Oracle Exalytics to uncover new revenue opportunities, enter new markets, develop new business models, and create new processes that can fundamentally transform an organization. Oracle Exalytics is empowering businesses at the workgroup, departmental, and enterprise level by delivering unprecedented analytical capabilities that can drive incremental revenues and margin growth, provide a competitive advantage, and lead to improved customer satisfaction.
For example, R.L. Polk & Co., a cloud-based automotive data and marketing solutions leader, sees Oracle Exalytics as contributing to top-line revenue growth by providing more customers with more data. "We see the Oracle Exalytics high performance allowing our customers to interact and fly through data without wait times," says Kelly Garcia, vice president, Production & Business Intelligence Technologies at Polk. Oracle Exalytics is helping R.L. Polk transform its business by providing speed-of-thought insights on billions of rows of worldwide automobile market data.
Key Energy Services, a U.S.-based energy services provider, sees Oracle Exalytics dramatically cutting down on analysis time and providing the ability to slice and dice through billions of rows of sensor data at the speed of thought. "With Oracle Exalytics, we were able to develop a performance dashboard in one-fifth the estimated time it would take to develop otherwise," says John Hood, vice president and chief information officer of Key Energy Services.
Savvis, Inc., a public cloud infrastructure service provider, sees the Oracle Exalytics scalability, performance, and visualization capabilities delivering much-needed cost savings by standardizing on an enterprise analytics platform versus one-off departmental solutions.
Finally, NTT Data IntelliLink deployed Oracle Exalytics with Oracle Exadata and achieved extremely high levels of maintainability and performance. NTT was able to analyze the big data loads much faster, which resulted in high-speed business analysis and fast decision-making.
Oracle Exalytics In-Memory Machine is an integrated hardware and software solution that is designed to deliver extreme performance for business intelligence and enterprise performance management applications.
Oracle Exalytics features a single 3U rack-mountable server powered by four Intel Xeon processor E7-4800 product family processors, each featuring up to 10 CPU cores for a total of 40 CPU cores. It features 1 TB of RAM to manage analytics data in-memory. Oracle Exalytics includes a high-performance, direct-attached storage system that includes a high-performance RAID HBA and 3.6 TB of raw disk capacity. To enable fast network interconnects, Oracle Exalytics features two quad-data rate (QDR) 40 GB/sec InfiniBand ports, and two 10 GB/sec and four 1 GB/sec Ethernet ports.
This abundance of compute capacity, memory, and fast networking options is matched by analytic software to deliver extreme analytics. Oracle Exalytics comes preinstalled with Oracle Business Intelligence foundation, Oracle TimesTen In-Memory Database, and the Adaptive In-Memory Cache Management feature.
Oracle Business Intelligence foundation delivers complete business intelligence functionality including enterprise reporting, dashboards, ad hoc query and analysis, online analytical processing (OLAP), scorecards, and predictive analysis. The Oracle Business Intelligence foundation feature set is complemented by a rich end user experience that includes advanced visualizations, collaboration, alerts, notifications, search, and mobile access. It also includes Oracle Essbase, the industry-leading, multidimensional OLAP server that provides a comprehensive environment for effectively developing custom analytic and enterprise performance management applications.
Oracle TimesTen In-Memory Database is a memory-optimized, full-featured relational database with persistence. It stores all its data in memory-optimized data structures and supports query algorithms specifically designed for in-memory processing. Using the familiar SQL programming interfaces, Oracle TimesTen In-Memory Database provides real-time data management that delivers exceptionally fast response times and very high throughput for a variety of workloads.
Most business intelligence deployments have workload patterns that focus on a specific collection of "hot" data from the enterprise data warehouse. In the past, the most efficient way to provide subsecond response times was by identifying and creating a data mart for rapid access to the most relevant hot data. But creating a data mart for query acceleration often requires expensive, error-prone manual research to determine which subject areas or cubes to bring into memory. In contrast, Oracle Exalytics features a summary advisor that automatically caches the most frequently accessed hot data in-memory based on system usage patterns. For data not available in-memory, Oracle Exalytics transparently ships the queries to underlying data sources. This intelligent management of the 1 TB of system memory ensures that resources are wisely used to hold the most frequently accessed data in-memory. This adaptive memory management dramatically reduces the cost of implementing and maintaining business intelligence deployments.
Both Oracle Business Intelligence foundation and Oracle TimesTen In-Memory Database for Exalytics have been specifically enhanced for analytical processing at in-memory speeds and can take advantage of large memory, processors, concurrency, storage, networking, operating system, kernel, and system configuration of the Oracle Exalytics hardware.
In fact, key operational elements of the Oracle Exalytics solution have been greatly enhanced through optimizations that take advantage of the Intel Xeon processor E7-4800 product family. With the Oracle TimesTen In-Memory Database for Exalytics solution that works on data within the processor's Integrated Memory Management unit, dramatic latency reductions for query processing, data volume management, and analytical conclusions have been achieved. This kind of optimization results in better query responsiveness, higher user scalability, and markedly lower TCO compared to standalone software.
Over the years, storage performance has lagged far behind processor performance. As a result, data access latencies have become a common source of performance bottlenecks in today's transactional and analytics applications. Major performance gains are also provided by the new generation of in-memory database solutions. Maintaining data and key components of the application software in system memory dramatically reduces the need for disk access and enables order-of-magnitude performance improvements. The Intel Xeon processor E7-8800/4800/2800 product families have been designed to unleash the benefits of in-memory systems such as Oracle Exalytics through enhancements for data visualization, analysis, and discovery. With in-database analytics, queries are performed directly on data residing within the database. This eliminates the delays associated with moving data from a data warehouse into a separate analytics system. It can also help to reduce infrastructure and administrative costs and complexity.
Oracle Exalytics includes two in-memory analytics engines to manage the data in-memory for speed-of-thought analytics: Oracle TimesTen In-Memory Database for Exalytics and Oracle Essbase. These two data management engines are leveraged in the following four techniques to provide high-performance, in-memory analytics for a wide variety of business intelligence usage scenarios at the workgroup, departmental, and enterprise scale:
All of these in-memory methods are assured by the reliability, availability, and serviceability (RAS) features within the Intel Xeon processor E7-8800/4800/2800 product families. These RAS features natively operate in hardware and include supporting software optimizations to ensure that low-latency Oracle TimesTen In-Memory Database for Exalytics constructs are robust in nature and provide guaranteed actions on data and data insights.
The Intel Xeon processor E7-8800/4800/2800 product families support up to 16 dual inline memory modules (DIMMs) per socket at 32 GB each for a memory capacity up to 2 TB in a four-socket system. Support for larger DIMM sizes allows more data to be analyzed in-memory.
Many enterprise business intelligence implementations, including prepackaged business intelligence applications provided by Oracle, may be able to fit entirely in-memory. In addition, it is possible to expand memory capacity through clustering or compression. When required, the Aggregate Persistence feature in Oracle Business Intelligence Server can be used to replicate the entire data warehouse in the Oracle TimesTen In-Memory Database. Incremental refreshes using double buffering can ensure that the in-memory data is completely in sync with the underlying data sources. This data replication mechanism is different from physical replication in that it is an extension of the Oracle Business Intelligence Server's logical model. This mechanism allows for in-memory analytics for all use cases, including ad hoc analysis and interactive dashboards.
This technique relies on the adaptive in-memory cache management software to analyze the workload patterns and caching the commonly accessed "hot" data from underlying data sources in-memory. In this case, the subsecond interactivity is provided by identifying and creating an adaptive data mart for the relevant hot data. This data mart adapts to the usage pattern and intelligently manages the 1 TB of RAM to move the data objects in and out of memory based on frequency of use. For example, the adaptive in-memory data mart may primarily consist of financial data during the close of a fiscal quarter and at other times it may include other data that's most relevant to business in-memory.
The Oracle Exalytics Intelligent Results Cache feature in Oracle Exalytics is a completely reusable in-memory cache that is populated with the results of previous logical queries generated by the Oracle Business Intelligence Server. In addition to providing data for repeated queries, any result set in the result cache is treated as a logical table that is able to satisfy any other queries that relate to a subset of the cached data. For the best query acceleration, Oracle Exalytics provides tools that can analyze usage and identify and automate the preseeding of result caches. This preseeding ensures instant responsiveness for queries at runtime.
Oracle Essbase with its in-memory optimizations for Oracle Exalytics provides yet another dimension for accelerating queries on specific subject areas. In contrast to other in-memory techniques, these cubes may be writable as well. As mentioned earlier, the Oracle Business Intelligence Server uses its Aggregate Persistence feature to create cubes out of data extracted from the semantic layer. This supports advanced scenario modeling and what-if analysis, and provides an unprecedented and seamless modeling and reporting framework.
Tests with customer data have shown query response times reduced by 20x to 100x, and throughput increases of up to 5x. Customers reported cutting down the planning and forecasting cycles by as much as 4x to 6x. In many cases the customers were able to run their entire data warehouses in-memory resulting in dramatic performance improvements.
Oracle TimesTen In-Memory Database for Exalytics supports an advanced columnar compression capability that reduces the memory footprint for in-memory data. Compression ratios of up to 5x are achievable and help to greatly expand in-memory capacity. Analytic algorithms that are designed to operate directly on compressed data further speed the in-memory analytics queries. In addition, multiple Oracle Exalytics machines can be clustered together to expand the in-memory capacity.
Oracle Exalytics provides an extremely powerful end user experience that includes best-in-class reporting, dashboards, ad hoc queries, OLAP, and scorecards. Knowledge workers can access information using the format and delivery channel that's best suited to their role, preferences, and experience level. Information can be delivered through dashboards, scorecards, and reports; within enterprise portals and collaboration workspaces, business applications, and Microsoft Office tools; and on mobile devices.
Oracle Exalytics includes a number of user-interface enhancements that promote interactivity and responsiveness by tapping into the system's powerful compute capacity and abundant memory. Visual analysis without limits is enabled by intelligently caching the "hottest" data in-memory using the resident 1 TB of RAM. For data that's not available in-memory, Oracle Exalytics transparently ships those queries to underlying data sources. Features like auto-complete prompt fields, cascading prompts, and contextual right-click interactions make it easier for the end user to analyze data and capture insights at the speed of thought. A dynamic user interface manages large amounts of data under the covers—while rich visualization capabilities present business information in meaningful and compelling ways. A wide range of interactive chart types, maps, and tabular formats are available to help visualize query results.
Oracle Exalytics features new microcharts and multipanel trellis charts to visualize dense, multidimensional, and multipage data on a single screen. The multipanel trellis charts are particularly effective at visualizing trends by displaying multiple visualizations across a common-axis scale for easy comparison.
The visualization and interactivity enabled by Oracle Exalytics is also available on supported Apple iOS mobile devices, including the Apple iPad and iPhone. With out-of-the-box support for most business intelligence and enterprise performance management applications, Oracle Exalytics enables organizations to deploy mobile analytical applications to a larger population of mobile users.
Oracle's software stack is the foundation for a powerful line of engineered systems that will help you quickly find new insights and unlock the value in any data. And that includes big data—vast quantities of unstructured information that's challenging organizations to find an efficient and timely way to acquire, organize, and manage it. Oracle's pre-integrated engineered systems address these challenges, reducing the cost and complexity that can plague IT infrastructures while increasing productivity and performance. Because we design the complete system, we can innovate and optimize at every layer of the stack to improve overall data management, simplify data center operations, drive down costs, and accelerate business innovation.
Business analytics projects initiated by a business unit often require just-in-time deployments that can yield a fast return on investment (ROI). And it is the IT department that has to deliver solutions that are reliable, available, maintainable, and secure to meet those needs. To deploy these solutions successfully, the business analytics servers and data warehouse components must work in combination to deliver real business value and robustness. Designs that simply focus on business analytics servers and in-memory solutions can miss the significance of establishing persistent, trusted data—an organization's historic database of record—in a data warehouse.
Similarly, if you focus on just the data warehouse you can miss the importance of putting the right data discovery infrastructure in the hands of business analysts.
Of course, the best-performing infrastructure includes a combination of optimized middle-tier business analytics server solutions and data warehouse solutions and provides short-term benefits and long-term viability.
Today, IT infrastructures are evolving to include engineered systems that can host both back-end data warehouses (such as the Oracle Exadata Database Machine) and middle-tier business analytics servers (Oracle Exalytics In-Memory Machine). The back-end solutions speed query resolution and in-database analytics performance by providing balanced servers with adequate throughput and database optimization in storage. The middle-tier solutions speed query performance by providing large memory areas that are used as a cache for frequently accessed data.
The following diagram illustrates key software supported in the Oracle Exadata Database Machine hosting a data warehouse and the Oracle Exalytics In-Memory Machine hosting the business analytics components.
The following diagram illustrates a typical topology of Oracle Exadata and Oracle Exalytics working together. The InfiniBand connection between Oracle Exalytics In-Memory Machine and Oracle Exadata Database Machine enables faster data movement than is possible using a 40 Gb Ethernet connection. Analytics can be performed where it makes the most sense when solving a business problem—in the middle tier or in the database. The platforms share common implementation services and the same single Oracle point of support. It's a combination that can't be beat.
You're investing in collecting big data because you know it holds valuable information about your customers, your competition, and about market and industry trends. By combining insights gleaned from big data with the traditional transactional data already available in your business, you can capitalize on previously hidden relationships and get a real competitive advantage. There are four key characteristics that define big data and provide potential challenges: volume, velocity, variety and value. Volume and velocity aren't necessarily new problems for IT managers; these issues are just amplified today. The distinguishing characteristics of big data that do create new problems are the variety and low density value of the data. Big data comes in many different formats that go beyond the traditional transactional data formats. It is also typically very low density; a single observation on its own doesn't have a lot of value. However, when this data is aggregated and analyzed, meaningful trends can be identified.
Oracle's powerful line of engineered systems form a big data platform that can help you quickly unlock the value in big data. These systems enable organizations to deploy big data solutions as a complement to operational systems, data warehousing, analytics, and business intelligence processing. Engineered systems are pre-integrated, so they are easier to deploy and support, and they deliver optimized performance. They can be deployed alone or alongside existing infrastructure.
In addition to adding volume to existing data warehouses faster, the data explosion includes massive amounts of new, varied information generated by social media networks, websites, streaming devices, as well as enterprise content such as documents, emails, and customer calls and chats. This diverse information contains critical insights about your business. But separating the signal from the noise to unlock new insights is a challenge for many organizations.
Oracle Endeca Information Discovery, certified to run on Oracle Exalytics, provides advanced exploration and analysis of various unstructured data like social media, product reviews, and enterprise documents across diverse sources. Oracle Endeca Information Discovery's intuitive interface empowers business users to easily search, navigate, and explore big data to find new opportunities like improving products based on better understanding customer needs. Combined, Oracle Endeca Information Discovery and Oracle Exalytics provide the optimal environment to derive value from big data.
Operating in today's big data world, you need the right tools to capture and organize a wide variety of data types from different sources, and to be able to easily analyze it within the context of all your enterprise data. Oracle's engineered systems and complementary software provide an end-to-end value chain to help you unlock the value of any data, and get the most from your data center.
Oracle offers a lifecycle of services for Oracle Exalytics. Oracle Exalytics Start-Up Pack provides an integrated solution to help plan, install and configure Oracle Exalytics for a fast and effective deployment. For further deployment activities to accelerate your return on investment, Oracle Consulting offers upgrade and implementation services as well as migration services, including the Oracle Migration Factory based on tools and best practices.
To help maximize performance and availability, Oracle Advanced Customer Support Services delivers tailored, mission-critical support services with a solution support center and advanced monitoring and resolution services for Oracle Exalytics. Oracle Premier Support offers essential support services including 24/7 technical assistance, powerful proactive support resources, and product updates.