EXASOL extends Hadoop integration to include all structured and unstructured data
EXAPowerlytics Hadoop integration service for rapid SQL and Map/Reduce analyses
Nuremberg, 09.04.2013 – EXASOL, the manufacturer of the EXASolution high-performance in-memory database, has extended its integration to Hadoop, the open-source platform for the distributed processing of extremely large quantities of data (“big data”). With the EXAPowerlytics Hadoop Integration Service, companies are now able to integrate large quantities of unstructured data from Hadoop into the EXASOL data warehouse. This can be linked with the existing company data such as customer profiles or financial ratios and subsequently analysed. With the new and significantly enhanced functions in terms of compatibility, productivity and manageability, EXAPowerlytics Hadoop Integration Service unleashes the full power of EXASolution and Hadoop. Companies are therefore able to boost their return on invest. They gain more in-depth knowledge of their business processes sooner – and that with minimal effort and cost.
Simple interaction between Hadoop and EXASolution
With Hadoop, large quantities of data can be analysed using Map/Reduce programs and algorithms in parallel on computer clusters. The log files of several thousand web servers are searched frequently to summarise data for reports. Companies can also use Hadoop to compile and convert data or to analyse a large number of images, texts and social networks, for example. With EXAPowerlytics, EXASOL enables the incorporation of external services in popular programming languages into the parallel SQL processing of EXASolution. The Hadoop Integration Service constitutes a special solution that assumes communication between EXASolution and Hadoop. Transparent access to Hadoop is possible at all times. Given that EXASolution itself has Map/Reduce capabilities, the analysis work can even flexibly be distributed between Hadoop and EXASolution. This means that data warehouse is combined with a Hadoop system in the simplest, most efficient manner.
“Our unique selling point primarily lies in the fact that we are able to directly influence the Map/Reduce algorithms that run in Hadoop,” explains Sebastian Klenk, Product Manager at EXASOL. “While the manufacturers of comparable databases can only access the Hadoop Distributed File System (HDFS) or import the results from entire Hadoop jobs, EXASolution enables the flexible division of certain partial jobs, such as the Map and Reduce step, between Hadoop and EXASolution. The advantages are clear: selected tasks can be delegated to Hadoop, and the data analyses continue to run in parallel and offer extremely high performance in EXASolution.”
An overview of the advantages of the EXAPowerlytics Hadoop Integration Service:
• Analysis jobs can be distributed flexibly between Hadoop and EXASolution
• Fastest possible linking of the data warehouse data with Hadoop data
• Native interlocking with SQL
• High-performance data transfer between Hadoop and EXASolution
“Nowadays, industries such as retail, e-commerce, telecommunications and financial services require a better understanding of customer behaviour,” says Steffen Weissbarth, CEO at EXASOL AG. “Companies use Hadoop to evaluate unstructured data from social media, clickstreams, call centre activities or other customer interactions. If companies or their specially-trained experts (data scientists) go one step further though, and link these digital data flows with the structured company data via EXAPowerlytics, they achieve an entirely new quality for their analyses, as they become considerably more meaningful,” Weissbarth continues.
The actual added value for the customers therefore lies exactly where unstructured data outside of the database comes together with structured data in the relational databases. EXASOL positions itself particularly through its long-standing experience as a noted expert in this field.
Making effective use of big data (value data)
To be able to analyse value data, large quantities of data must be loaded rapidly so that the systems do not come to a standstill. The high-performance loading of data is a capability that EXASolution is able to provide, enabling value data to be made available more rapidly than ever before, and can be used immediately to make decisions. Traditional database systems require extensive, time-consuming configuration and optimisation to be able to deliver reports and analyses in an acceptable format, or special hardware platforms whose scaling can be difficult or costly given changing business requirements. Through a column-based massively parallel processing (MPP) architecture in addition to data compression and indexing technologies, EXASolution can manage countless analysis queries dynamically in parallel. Hence thousands of concurrent users, user-driven reports and ad hoc queries as well as data-driven processes throughout the entire company can simultaneously be supported easily and cost-effectively.
EXASOL AG, based in Nuremberg, develops and markets the high-performance database EXASolution, which is based on in-memory technology and was designed specifically for data warehouse applications and business intelligence solutions. This also allows extremely large volumes of data to be analysed and evaluated within the shortest of times. Thanks to the high performance and low administrative maintenance, EXASolution not only supports businesses with valuable decision-making bases from their data, but also reduces the total cost of ownership. In April 2011, Gartner named EXASOL AG a “Cool Vendor” in the category of “Data Management and Integration 2011”. In 2012 and 2013, the solution was included in the Magic Quadrant for Data Warehouse Database Management Systems.