Data blog — PostgreSQL RSS

Table partitioning in relational databases

I would like to write shortly about table partitioning in relational databases. Table partitioning is basically dividing your data in a table into horizontal chunks, that can be (depending on the DB technology you use) indexed separately and stored on different disks. This allows you to address certain performance issues, if a table is large, and there are many inserts into it, and there is a requirement of providing reports on the data in this table. That table might be for example a transaction registry from your retail network. Partitioning allows separate 'read only' partitions from the active partitions. For example if you insert a lot of transactions to your registry table, they usually have a timestamp associated with them. You...

Continue reading

Ranking of database engines

So it is January again and it is time for various trend reports and market share. Portal publishes, among other interesting content, periodic reports showing popularity of DBMS engines on the market. There is a number of factors taken into account, best described here. It is important to know that the report does not show the relative share of particular DBMS on the market, but rather its popularity among developers really.   What is striking though, that there are three leading technologies in terms of popularity - Oracle, MySQL and SQL Server. They are quite close to each other and quite apart from the fourth one - PostgreSQL. When you look at the historical trends chart, you can see that...

Continue reading