MySQL server logical architecture diagram is as follows:
Currently working with version 5.5, version 5.5 starts MySQL to start with InnoDB as the default storage engine, and InnoDB tables are based on clustered indexes.
MySQL Storage Engine lock management is very important, on a given resource, the less the amount of locked data, the higher the system concurrency, as long as there is no conflict between each other.
Shared locks are also called read locks, and exclusive locks are also called write locks.
Characteristics of a transaction acid,a atomicity (atomicity) C conformance (consistency) I isolation (isolation) d persistence (durability)
The isolation level of the transaction is: 1, read UNCOMMITTED unread (will produce dirty read, generally do not) 2, Read Committed commit reads (most database default) 3, repeatable read repeatable read (mysql default) 4, Serializable serializable (usually not used)
InnoDB handles the deadlock method and rolls back the transaction that holds the least row-level exclusive lock.
InnoDB is a transactional table, and MyISAM is a non-transactional table.
InnoDB 1, support transaction 2, support hot backup 3, crash recovery fast
High performance MySQL notes (Chapter One, MySQL architecture and history)