Time-series DBMS supports processing of large-scale sensor data.
Core technologies in smart X fields such as smart factories
v6.0, disk compression ratio increased by 3 times compared to previous versions


A time series DBMS is a special purpose DBMS for processing time series data generated at regular time intervals.

It is emerging as a core technology in smart X fields such as smart factories, smart cities, and smart grids, as it enables processing of hundreds of thousands to millions of sensor data per second, which was impossible with existing RDBMS and big data solutions.

On the 24th, Markbase announced 'MACHBASE v6.0', a time-series DBMS that supports real-time processing of ultra-large volumes of sensor data generated from IIoT devices.

In November of last year, Markbase ranked first in the TPCx-IoT field test conducted by TPC, a global accredited certification body, and was listed as an international standard. Instead of omitting the functions of existing RDBMS such as transactions and data changes, it adopted tag (sensor) and time-based index configuration and high-efficiency compression technology for real-time input and retrieval of large amounts of data reaching millions of records per second.

Kim Seong-jin, CEO of Markbase, said, “Until recently, it was not uncommon for smart factories to process tens to hundreds of thousands of sensor data per second.” He added, “This year, companies that require processing more than 70 million data per second have emerged, and the amount of sensor data that needs to be processed in the IIoT environment is exploding.”

The most notable feature improvement in v6.0, which was announced with the goal of storing, analyzing, and utilizing up to 100 million data records per second, is the Tag Table compression function to minimize disk usage, which is directly related to performance and cost. To improve price-to-performance ratio, it is important to maximize the number of transactions per second and minimize disk usage. By improving tag table compression efficiency, disk compression is possible up to three times compared to existing methods.

A new incremental backup method has also been added, allowing backups to be made only for data that has occurred since the last backup, reducing backup time and storage space. This can be applied to the entire database, tags, logs, and tables.

To ensure customer-required DBMS availability, we've also added multi-connection support for automatic failover in clustered environments. If the connection to the primary server is lost due to various reasons, including server failure, you can connect to a pre-configured server and restore the status of any queries being processed on the previous server.