The feature of IoT is that large amounts of data should be generated and responded in real time. What it needs is sensors and data processing solutions. Data processing solutions are very important to extract meaning from data and to respond quickly. In particular, a DBMS in the IoT era for processing a large amount of data in real time and in a time series are required.

Infiniflux (CEO, Kim Song-jin, photo) is an IoT DBMS company which has developed the world``s fastest IoT Big Data Real Time Series DBMS solution. It fused a search engine to make ``Infiniflux`` a ``monster`` based on the DB engine technology. Currently Infiniflux 3.0 has the industry``s best collection, loading, compression storage, search and analysis capabilities. It will release a cluster version that greatly enhances scalability in the second half of the year.

Machine data, log data and various sensor data are generated in IoT. IoT devices will be in excess of 50 billion in 2020. Prior to the introduction of IOT and Big Data, it was common to store big data in files and not use them. There was no solution to handle vast amounts of data and the perception of big data values was low. However, with the opening of the IoT, Big Data became a jewel-bearing mine, and the situation changed as Hadoop technology became available to handle big data storage.

However, both Hadoop and existing DBMSs have their limitations in real-time data processing at very high capacity. If it was good at storing and compressing, it lost its wits, and if it was good at searching, it was behind in its analysis. Infiniflux has developed a real-time time series large-capacity DBMS by combining traditional DB engine and big data search engine for four years since its establishment in 2013 in order to overcome these shortcomings. This is the grit of Kim Sung-jin, CTO and CEO of Altibase, who led the industry in the initial domestic DBMS solution industry.

President Kim said, ‘It can see the advantages and disadvantages of the existing Big Data solutions and is able to store large amounts of data, but the rapid response, scalability and easy operability of queries will be a competitive advantage for IOT DBMS. When anyone moves easily with a mouse, anyone was able to develop a product with so called slogan ``Super Easy``, which can find the ``number of business`` using Big Data DBMS."

In the meantime, Infiniflux set its own key value. The search performance comparable to Splunk along with low-cost expansion comparable to Hadoop, and compression storage capacity of Cassandra has been benchmarked. First, the generated data is stored on a columnar basis optimized for OLAP. Columns are tens of times faster than row-based. It applied a technology that can compress data up to tens of times in the logical physical layer. This is essential to minimize system load and improve data entry and retrieval performance.

LSM (Log Structured Merge) indexing, which is suitable for processing large amounts of data, has been applied to instantly analyze billions of data in data analysis. It takes only 0.5 seconds to retrieve one out of a billion even in a typical server environment. In addition a mount technique that allows users to instantly retrieve backed up data without loading data. Infiniflux also has a Web-based visualization capability. The general staff realized ``super easy`` that they could use Big Data Lane with a mouse click.

Infiniflux has been selected as a partner of ARTIK, the IoT platform of Samsung Electronics, in January, recognizing the difference in real-time time series analysis performance. It is equipped with an articulated IoT module and it is possible to enter the world market.

Infiniflux is now installed in the firewall of domestic leading security company S, which supports large-capacity log processing, management, search and reporting. Company F and real-time firewall logs with 4.5 billion events a day and more than 1TB events. In addition, it has a partnership with more than 20 partners to obtain a reference. The company established a US sales office in Silicon Valley last October. Infiniflux plans to move its headquarters to Silicon Valley and Seoul will perform R & D functions. Mr. Kim stated, ‘we would like to open up the cluster-based Infiniflux in the second half of the year and move the headquarters to Silicon Valley in order to compete in the global market. We would like to write a success story that will grow into a global company as a DBMS company.

By Lee Gyu hwan Editorial reporter david@


[ copyright ⓒ The Digitaltimes ]