The Computer Revolution/Databases/Database Classifications
Database Classifications: Single-User vs. Multi-user Single-user databases are located on a single computer and are designed to be accessed by one user. Multi-user database systems may try to access and modify the same data at the same time. Database locking software takes place to prevent data changing at the same time. Client-server vs. N-tier Client-server consists of servers that supply resources to other computers which function as client devices. N-tier databases work slightly differently. While client-server database systems just have two parts, N-tier has at least one middle client (programs used with database, or middleware). Centralized vs. Distributed database system Centralized database systems are all located on a single computer, while distributed database systems have the data divided among several computers connected via a network. Disk-based vs. In-memory Disk-based databases store data on hard drives or removable disks. In-memory databases store data in the main memory of the computer.
In-Memory Database
[edit | edit source]An in-memory database (IMDB, also known as Main Memory Database, or MMDB) stores all the data in the RAM of the computer as opposed to a hard drive. The advantage of an IMDB is faster performance (recall). The disadvantage is the potential loss of data in a power outage. Performing regular backups of IMDB’s is critical to ensure continuation of service if power is lost. Other methods of minimizing data loss include: using transaction logs, using non-volatile RAM, and complete database replication .