Client
Hewlett Packard, an American multinational information technology company headquartered in Palo Alto, California
Business Need
Hewlett Packard required a demonstration of the Big Data capabilities of their “HP Helion” Cloud. HP also wanted to give customers the ability to build and manage their own Big Data solutions in a few simple clicks. The Akvelon team was selected to tackle this problem and our team proposed multiple options to fulfill HP’s requirements. One of our options was selected and this option included automated private cloud deployment of Hadoop cluster and demonstrating data analysis. HP contracted Akvelon for design and implementation of entire solution from start to finish.
Solution
Akvelon developed an automated approach for private cloud Hadoop cluster deployment, primarily based on Ansible technology. Our team developed a sample data generator using node.js, capable of producing huge amounts of JSON data with pre-set trends, and implemented gathering generated data to HDFS using Flume. We then implemented data analysis using Pig script, which detected predefined trends in generated data. In addition, documentation and video guides were prepared for all parts of solution.
Benefits and Results
Akvelon successfully delivered the project on time for HP and provided additional support for HP by answering detailed technical questions.
HP received:
•Interesting and reproducible examples of HP Helion usage for Big Data
•Well-documented and maintainable Ansible scripts for Hadoop cluster deployment
•Data analysis examples for trend discovery using Pig scripts in Hadoop cluster
•Video material with description of every step of solution
Technology Used
HP Helion cloud, Ansible, Hadoo, HDFS, Flume, Pig, node.js
