What is hybrid partitioned table?

Altibase is the world’s first company to develop hybrid partitioned table.

A hybrid partitioned table is a table that is partitioned in a way that data can be effectively and efficiently segregated and stored in corresponding storage mediums according to the temperature of the data in a hybrid database. The purpose is to allow for performance enhancement particularly when a table has to process large data sets.

A hybrid database refers to a database that combines an in-memory database and a disk-resident database in a single unified engine. In the hybrid architecture, data can be stored and manipulated in the main memory, on disk or a combination of both. When you have frequently used (hot) data and need more speed, you store the data in memory and have an order of magnitude performance enhancement. When you have huge rarely accessed historical (cold) data, you store the data on disk and enjoy the low cost and economical storage scalability of traditional hard disks.

However, in many cases, hot and cold data is mixed in a large table. If you need to create multiple tables to handle data of various attributes originated from one table, you would have to deal with complex process logic, which would increase the complexity of programming and consequently impede performance.

A hybrid partitioned table is to cope with this problem in a hybrid database when hot data and cold data is mixed in one large table. A hybrid partitioned table is a method of logically grouping a memory table and a disk table into one table where data is sorted out and stored in memory partitions or disk partitions according to the user specification for partitions of a table.

By effectively processing hot data and cold data in one logical table, it provides a huge advantage in the form of efficacy and efficiency in a writing program with a net result of high performance enhancement.