|MySQL 5.6 - more speed and scalability|
|Written by Kay Ewbank|
|Tuesday, 12 April 2011|
An updated version of MySQL has been released promising more speed and scalability.
MySQL 5.6 is still essentially a beta release, part of the MySQL method whereby updated versions of the software are released as soon as any significant element is stable enough to test.
The announcement of the ‘milestone release’ was made at the MySQL conference in Santa Clara, and according to Oracle will allow users to test a number of features designed to improve performance. Oracle’s vice president of MySQL engineering, Tom Ulin, said:
The new release is also more flexible and compatible with cloud and web-based applications.
The improvements to performance have been gained by improving the query optimizer, improving InnoDB for higher transactional throughput, and handling partitioning better to make it easier to query very large tables. NoSQL style memcached APIs are also included.
The optimizer improvements are probably the most important change. An index condition pushdown means that for queries including WHERE clauses, more of the processing takes place in the storage engine, so instead of fetching an entire row that is then evaluated against the WHERE clause, only matching results are returned to minimize I/O traffic.
Another performance improvement comes from the inclusion of multi-range reads which deal better with secondary index lookups by scanning the index ranges in the query and sorts the disk blocks so the data can be returned using sequential I/O requests.
The InnoDB improvements come mainly from improved accuracy and better persistence across server restarts of the InnoDB index statistics. InnoDB is the default table type for MySQL. The index statistics are used by the optimizer to work out which indexes to use in a query and can have a significant effect on query performance.
The inclusion of NoSQL access methods means your web services can now directly access the InnoDB storage engine without transformations to SQL, ensuring low latency and high throughput for read/write queries.
SQL Clearly Explained, 3rd Ed (Book Review)
|Last Updated ( Tuesday, 12 April 2011 )|