This paper discusses some of the more salient aspects of slavery, particularly those found within this book that pertain to Virginia and its role during the civil war and as a slave state in general. Chattel slavery had a decimating effect upon America as a country on the whole. It produced a plethora of perverse relationships between people.
¶ … Gabriel's Rebellion: The Virginia Slave Conspiracies 180
There seems to be a growing consensus of analysts and product reviewers who believe that the most efficient form of BI for Big Data involves Hadoop. The increasing number of software solutions offered by a host of vendors in the last six months attests to this fact. This claims is further bolstered by the fact that there are several aspects of Hadoop that make it ideal for Big Data -- its virtually unlimited scalability, the real-time speed in which it grants access, and its cost efficiency (it's an open source platform that uses commodity hardware). Factor in the amount of variegated data that it can accommodate (which makes it desirable for little data projects as well) and the fact that it has its own system of analytics, and its popularity becomes understandable. Support for Hadoop is increasing, a fact which many NoSQL Big Data platforms cannot claim.
Although Hadoop has its own analytics, much of its BI value lies in its integration with existing software. Its components are powered by batch processing technologies which reduce the efficiency of its data storehouse, Hive, and its analytics. These technologies -- principally MapReduce and Hadoop Distributed Filing System -- are more rigid and slower than traditional BI, making it difficult to analyze data expediently. It's usually more advantageous to use external BI solutions that integrate with Hadoop either with or without Hive.
With the former option, BI products simply accesses data in Hive through their cache and use their own tools to analyze it without involving Hive's ponderous querying system, HiveQL. In addition to expediting the analytics process, Hive-based BI products can utilize its standards based approach to position reports within various distributions of Hadoop. These solutions offer all of the convenience associated with traditional BI including tools for visualization, analysis, forecasting and more, although they frequently aren't as fast as those that don't involve Hive.
Hadoop Without Hive: Real-Time
Users can optimize Big Data's value by integrating BI products with Hadoop without Hive. This is partly due to the fact that connectors for Hadoop's filing system are more commonplace than those for Hive -- meaning conventional BI tools can utilize the former without the latter. Users can also find more support for Hadoop than for Hive. Additionally, it is possible for data warehouses to store a finite amount of data from Hadoop without Hive.
But the principle benefit of using Hadoop's big data scalability without going through Hive is the potential for data analysis in real-time. Using Hive to cache data or using a data warehouse to store Hadoop data only comes close to immediate accessibility. There are a number of BI tools that can query Hadoop as data comes in. These tools include both open source and proprietary options, and present data in a variety of forms that utilize spreadsheets, graphical user interfaces and conventional BI options. They allow users to visualize the entire spectrum of data relevant to a particular decision from a variety of sources -- typically a lot more so than the other options discussed -- for immediate analysis. Particularly competitive solutions provide plugs to specific CRM products, popular social media outlets, and mainframes.
The other competitive advantage regarding BI and analytic tools specifically designed for Big Data is its ease of use. Whereas traditional BI commonly requires IT personnel to generate queries through prolonged data mining processes for reporting and visualization purposes, a number of analytic tools targeting Big Data utilize languages that program queries automatically. These tools are designed for business professionals to gain faster access to data without waiting for techies to process their requests, enhancing the immediacy of these solutions. The potential of this aspect of Big Data analytics makes it possible for smaller organizations without extensive IT departments to utilize these tools while generating more queries faster than before while expanding BI's users to include laymen.
Cost vs. Value
For large scale operations, the cost of real time BI tools for Big Data can average between a low to medium six figures. Most solutions can scale prices and the amount of data processed to fit an organization's specific needs. Yet in reducing the volume of data, organizations are effectively decreasing Big Data's business value. Big Data's appeal includes the size and variation of data it processes and its ability to unlock and incorporate unstructured data. Large scale real-time BI tools designed for Big Data provide valuable insight into relationships between different types of data that are not apparent until they are streamed, queried and presented in a report or some form of visualization. The scalability of the analytics of these tools and their ability to process a multitude of data sources is where their real business value lies.
You’re 88% through this paper. Sign up to read the full paper.
Sign Up Now — Instant Access Already a member? Log inAlways verify citation format against your institution’s current style guide requirements.