SEOUL, KOREA — The most pressing hot-button issue in the IT industry is unassailably big data. As a matter of fact, it is the hottest issue across the entire industry spectrum. Big data gushing out of social networking services is expected to serve as the key to ramping up the competitiveness of companies. Big data sees no boundaries. In this era of big data, various data that had been left underutilized within companies has been thrown back into the spotlight.
In celebration of the big data era, Oracle announced that it will seek solid total data architecture that features the handing of formatted data with maximized performance and stability as well as large-volume unformatted data with data-specific analysis and management. In other words, based on Oracle’s database technology, Oracle will take the processing speed and stability of formatted data to new heights through Oracle Exadata (a database machine with support for both data warehousing and OLTP applications).
In July, 2012, Mark Hurd, President of Oracle Corporation, paid a visit to Korea and unveiled the “Oracle Optimized Data Center Strategy,” which Oracle will utilize to fast-track business innovations. This will be done by proposing new datacenter optimization methods that maximize simplicity and integration to live up to growing demands for big data and cloud computing.
However, since big data is very different from previous data in every aspect including the volume of data traffic and processing speed, support from bid data experts is badly needed. Under such circumstances, Edward J. Yoon, one of Korea’s leading software developers, has joined Oracle.
A Wake-up Call to Big Data
Edward J. Yoon said, “Though my interest in software development came unexpectedly during my younger years, I realized that I had a strong affinity for software development.” His interest in software came to fruition while he was working on the construction of an online fashion shopping mall for his sister. As a student, he dabbled in designing online shopping malls simply to make pocket money. Yet, by doing so, he found out that he was exceptionally adept at developing software. Mr. Yoon, who double majored in math and image processing in university, is a self-taught software developer.
As the online shopping malls he managed grew, problems piled up in proportion to growth in data such as sales items, merchandise information, and customer reviews. As he tackled problems one by one, he found out that all of the problems eventually boiled down to data management. In an effort to solve such problems, he turned his attention to information retrieval, distributed computing and data growth. This was the moment when Mr. Yoon began to wrestle with the issue of big data, which is a pressing issue these days. In 2005, Korea’s Internet penetration rate grew steadily while other nations with larger online populations started to gain a taste of big data such as information searches and distributed computing.
When Mr. Yoon attended an academic society meeting held in a foreign nation in 2007, he received hands-on experiences with big data. During the meeting, he saw other attendees browsing Facebook via their smart devices like the i-Phone and the iPod touch. He recalled that the sight of people going online via mobile gadgets came as a shock to him, who at that time carried a Motorola Razr phone.
Mr. Yoon Opens His Eyes to Open Source
After the meeting, Mr. Yoon took note of Apache Hadoop (an open-source software framework that supports data-intensive distributed applications, licensed under the Apache v2 license) and started to work as an Apache Software Foundation (ASF) committer (a developer that is given write access to the code repository). The membership-based ASF is the most active non-profit organizations incorporated in the US. Mr. Yoon is one of the few Asian committers; the majority of committers are European, American and Indian countries. The other Korean ASF committers include Trustin Lee (Twitter) and Chang-Shin Lee.
The ASF is run strictly under the principle of "meritocracy" (literally, government by merit). According to the degree of contribution, membership and the right to partake in ASF projects are granted. Mr. Yoon has been credited with catapulting Apache Hama (an open-source volunteer project under the auspices of ASF Incubator Project) into one of the top-level Apache projects. As of now, he serves as the Chair of the Apache Hama Project and an ASF member. Apache Hama is a pure BSP (Bulk Synchronous Parallel) computing framework on top of HDFS (Hadoop Distributed File System) for massive scientific computations such as matrix, graph, and network algorithms.
The ASF is comprised of member communities scattered around the globe. Software developed by members is either free or open source software. Software developed by members is distributed under Apache licenses. Apache projects are defined by collaborative, consensus-based development processes and practical, open-source software licenses. Each project is managed by expert groups consisting of volunteering members. The ASF hosts a large-scale conference annually to facilitate the sharing of technology and information related to Apache projects among members.
Apache Software Developer
Mr. Yoon mentioned, “Since I began making contributions to the ASF, I have been receiving job offers from major global IT companies such as Facebook and Google. Mr. Yoon, who has headed up the Apache Hama Project since 2007, is the creator of the project. Apache Hama, an open-source version of Google’s Pregel, has been developed to address the shortcomings of Apache Hadoop.
Apache Hama adopts the technology of message passing to process complex, iterative algorithms in a faster and more effective manner than MapReduce (a programming model for processing large data sets). In addition, since it provides a programming model that enables the parallel and distributed processing of algorithms (convertible to graphs), complex, parallel graph algorithms can be easily developed. As algorithm codes are divided among a number of distributed memory machines, faster results can be obtained.
Apache Hama, which is still in its infancy, can be put to good use for various purposes such as graph data, social network analysis, and DNA analysis. Apache Hadoop, the open-source equivalent of Google's MapReduce, is a distributed computing framework that is being developed by the Apache camp led by Yahoo. Amazon and Yahoo have applied Apache Hadoop to some of its commercial systems; Korea’s top search portals such as NHN (Naver) and Daum are also using it.
Working as a Mentor on “Superstar Software Developer K”
Mr. Yoon used to be a self-employed software developer who built online shopping malls on his own. Later, as he built up his reputation working as an Apache committer, NHN Corporation in 2007 hired him as a senior software engineer in charge of unformatted data processing. He worked for NHN for five years before he was recruited by Korea Telecom. Then, in July of this year, he entered Oracle to take charge of big data.
Mr. Yoon added, “My father, who spent his whole life working as a mechanic in the air force aviation science and mechanics fields, was a man with a free spirit. He is definitely the greatest influence on me.” Describing his father as a free-spirited, motorcycle-riding guitarist, Mr. Yoon said he had also dreamed of becoming a musician at some point in his life. Yet, Mr. Yoon knew that he was always a math lover to the core. He found that even the sight of guitar frets reminded him of geometric sequences. Mr. Yoon is a man who enjoys discovering the aesthetic of mathematics in the little things of everyday life.
Mr. Yoon is now increasingly being asked to give lectures on big data. He is currently serving as a mentor on “Superstar Software Developer K,” an open-source software promotion project launched by the National IT Industry Promotion Agency (NIPA). The future moves made by this famed software developer indeed merit due attention.