Altibase announces that it has recently developed a partitioned table with a view to greatly enhancing the overall performance in dealing with big data.
A hybrid partitioned table is a table that is partitioned in a way that data can be effectively and efficiently sorted out and stored in corresponding storage mediums according to the temperature of the data in a hybrid database. The purpose is to enhance performance particularly when a table has to process large sets of data.
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 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.
However, in many cases, hot and cold data are mixed in a large table. If you need to create multiple tables to handle data of various attributes, it will be difficult to manage and waste of storage space.
For example, if you want to make a small amount of hot data in a memory table and a large amount of cold data in a disk table, you would have to deal with complex process logic, which would increase the complexity of programming.
A hybrid partitioned table is to cope with this problem in a hybrid database when hot data and cold data are mixed in a 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 automatically sorted and stored in a memory or a disk partition according to the temperature of the data.
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 writing program with a net result of significant performance enhancement.
After 20 years’ experience as a closed-source database, Altibase is now open source, and that includes its cutting-edge sharding.