Toshiba’s Distributed Database GridDB® Now Features Scale-Out and Scale-Up combo for Petabyte-scale Data Management
December 3, 2019
Toshiba Digital Solutions Corporation
KAWASAKI―Toshiba Digital Solutions Corporation (Toshiba) today announced the general availability of GridDB 4.3, a purpose-built database for IoT and Big Data workloads. The latest version features petabyte*1-scale data management deployable even with a small number of servers.
In IoT systems, petabyte-scale data management is the norm due to the long-term accumulation of sensor data generated by connected devices. There are cases where building a large-scale cluster consisting of several dozens of servers are expensive and a smaller cluster is preferred instead. In this regard, increasing the maximum database size per server becomes essential.
By optimizing the internal data management structure of GridDB, the resource usage is greatly reduced and the maximum database size per server is increased. In addition, improvement in database buffer control, data placement and composite indexes paves the way to a better, faster, large-scale data processing performance. Petabyte-scale data management in the order of milliseconds with only a small number of servers is now possible using the scale-out*2 and scale-up*3 combo.
GridDB has been successfully implemented in various industry verticals, and is processing time-series data for critical systems in the energy, manufacturing and social infrastructure sectors. Toshiba will continue to enhance functions of GridDB, in support of digital transformation with IoT and Big Data analytics.
- *1 Petabyte: 1015 bytes
- *2 Scale-out: to increase processing capacity by adding more servers to a system
- *3 Scale-up: to increase processing capacity by adding resources such as CPUs or memory to an existing server
Toshiba developed GridDB entirely in-house, drawing on its extensive expertise in diverse industry verticals. The database combines the power of SQL and NoSQL to ensure efficient accumulation of massive volumes of time series data, and to deliver scalable performance. It features an IoT-oriented data model, in-memory architecture, excellent scalability and unwavering reliability—essential characteristics of IoT and Big Data database.
- * GridDB is a registered trademark of Toshiba Digital Solutions Corporation in Japan.