IBM PureData System for Hadoop

IBM PureData System for Hadoop - is optimized for accomplishment of the fast analysis and supports archive with a possibility of formation of requests. It contains the built-in accelerators of analytics for data from social networks, data analysis and the text and also includes the unique built-in means of archiving.

Inay SU (Inhi Suh) represents IBM PureData System

Description

Historically it developed - information is stored in the companies, as a rule, on several platforms. Data are stored in the unstructured, not optimized type that limits possibilities of extraction from them of the data necessary for decision making. A task of Big Data here - in an opportunity to collect such information from different input data, to structure it and to issue data for use in the analysis, decision making and work with means of predictive analytics.

The IBM PureSystem system, integrating in itself the concept of Big Data and the solution Apache Hadoop, it is created proceeding from these tasks.

The Apache Hadoop library makes the distributed processing of big data sets. Simple programming models under Hadoop are for this purpose used. Basic purpose of Hadoop - to provide process management of data processing on a set of servers and their synchronization, but only at the expense of the software, having removed management at the cluster, hardware level.

The IBM PureData for Hadoop system was created taking into account such approach regarding hardware and the software prepared for a cloud computing architecture. All advantages and features of Hadoop integrate with support and simple administration which PureData can offer.

For ensuring integration of Hadoop and this system integrated IBM InfoSphere BigInsights and the IBM System x servers, thus, the software for processing of big data sets is integrated in a complex, simple in terms of administration, and updates are made by IBM company for all computer system.

Built-in expert knowledge

• Acceleration of payback of data of large volume
• Deployment is carried out eight times quicker, than for individually created clusters
• The built-in opportunities of visualization facilitate obtaining valuable information
• Built-in accelerators of analytics

The simplified interaction

• Simplification of work with large volumes of data
• The single system console for complete administration of a system
• The fast and automated processing of updates
• Assembly is not required, readiness for data loading in several hours

Close integration

• Implementation of data of large volume of the corporate level
• Only for the integrated Hadoop system with the built-in instruments of archiving
• Higher security in comparison with open source applications
• The architecture providing high availability

The video presentation of IBM PureData System from Inay of SU (Inhi Suh) (engl.)