This chapter covers the MySQL architecture, locking and concurrency, and transactions. It also discusses how to select the right engine and looks at each of MySQL's storage engines in detail. (From the book High Performance MYSQL: Optimization, Backups, Replication and Load Balancing, by Jeremy Zawodny and Derek Balling, ISBN: 0596-003064, O'Reilly Media, 2004.)
MySQL provides in-memory Heap tables for applications in which you need incredibly fast access to data that either never changes or doesn’t need to persist after a restart. Using a Heap table means that a query can complete without even waiting for disk I/O. This makes sense for lookup or mapping tables, such as area code to city/state name, or for caching the results of periodically aggregated data.
Limitations
While Heap tables are very fast, they often don’t work well as replacements for disk-based tables. Until MySQL Version 4.1, Heap tables used only hash-based indexes rather than B-tree indexes (which MyISAM uses). Hash indexes are suited to only a subset of queries. The section “Heap Tables” in Chapter 4 covers this in more detail.
Berkeley DB (BDB) Tables
MySQL’s first transaction-safe storage engine, BDB is built on top of the Berkeley DB database library, which is now maintained and developed by Sleepycat Software. In fact, the original work to integrate the Berkeley DB technology with MySQL was performed jointly by MySQL AB and Sleepycat Software. Other than transactions, the BDB table handler’s other main feature is that it uses page-level locking to achieve higher concurrency than MyISAM tables.
Though BDB tables have been available in MySQL since Version 3.23, they haven’t proven very popular among users. Many users looking for transactions in MySQL were also looking for row-level locking or MVCC. Further dampening interest in BDB, by the time the BDB code had stabilized, word of InnoDB began to circulate. This prompted many users to hold out for the real thing and use MyISAM tables a bit longer.
If nothing else, the inclusion of BDB tables in MySQL served as a stepping stone in many ways. It prompted the MySQL developers to put the transaction-handling infrastructure into MySQL, while at the same time proving to the skeptics that MySQL wasn’t a toy.
If you've enjoyed what you've seen here, or to get more information, click on the "Buy the book!" graphic. Pick up a copy today!
/ 55 
