Metatron engine: Druid¶
The development of information and communications technology has been accompanied by a rapid increase in the amount of data generated, highlighting the importance of efficient data collection, management, and utilization. However, RDBMS-based legacy tools are unable to process mass amounts of multidimensional data. This has led to the emergence of new methodologies and solutions aimed at satisfying the demand for big data.
Metamarkets, a technology startup based in Silicon Valley, launched a column-oriented distributed data store known as Druid in 2011, and open sourced it in October 2012. Many companies have turned to Druid for their backend technology because it offers various advantages, including fast and efficient data processing.
As a B2C telecommunications service provider, SK Telecom recognized the need to effectively manage and analyze the vast amounts of network data generated by its users every minute. Metatron, an end-to-end business intelligence solution with Druid as the underlying engine, was thus developed and launched in 2016.
The following sections discuss the features of Druid that make it suitable for time-series data processing, and introduce how they were adapted and improved by SK Telecom for Metatron.
- Background of Druid development
- Druid features
- Druid cluster architecture
- Druid performance assessments
- Metatron powered by Druid