read comitted和repeatable read的不同之处

转自：http://stackoverflow.com/questions/4034976/difference-between-read-commit-and-repeatable-read

down voteaccepted

Read committed is an isolation level that guarantees that any data read wascommittedat the moment is read. It simply restricts the reader from seeing any intermediate, uncommitted, 'dirty' read. IT makes no promise whatsoever that if the transaction re-issues the read, will find theSamedata, data is free to change after it was read. Repeatable read is a higher isolation level, that in addition to the guarantees of the read committed level, it also guarantees that any data readcannot change, if the transaction reads the same data again, it will find the previously read data in place, unchanged, and available to read. The next isolation level, serializable, makes an even stronger guarantee: in addition to everything repeatable read guarantees, it also guarantees thatnonewdatacan be seen by a subsequent read. Say you have a table T with a column C with one row in it, say it has the value '1'. And consider you have a simple task like following: BEGIN TRANSACTION; SELECT * FROM T; WAITFOR DELAY '00:01:00' SELECT * FROM T; COMMIT; That is a simple task that issue two reads from table T, with a delay of 1 minute between them. under READ COMITTED, the second SELECT may returnanydata. A concurrent transaction may update the record, delete it, insert new records. The second select will always see thenewdata. under REPEATABLE READ the second SELECT is guaranteed to see the rows that has seen at first selectunchanged. New rows may be added by a concurrent transaction in that one minute, but the existing rows cannot be deleted nor changed. under SERIALIZABLE reads the second select is guaranteed to seeexactlythe same rows as the first. No row can change, nor deleted, nor new rows could be inserted by a concurrent transaction. If you follow the logic above you can quickly realize that SERIALIZABLE transactions, while they may make life easy for you, are alwayscompletely blockingevery possible concurrent operation, since they require than nobody can modify, delete nor insert any row. The default transaction isolationlevel of the .NetSystem.Transactionsscope is serializable, and this usually explains the abysmal performance that results. And finally, there is also the SNAPSHOT isolation level. SNAPSHOT isolation level makes the same guarantees as serializable, but not by requiring that no concurrent transaction can modify the data, but by making every reader see it's own version of the world (it's own 'snapshot'). This makes it very easy to program against, very scalable as it does not block concurrent updates, but of course it has a price, and the price is extra server resource consumption. Supplemental reads: Isolation Levels in the Database Engine Concurrency Effects Choosing Row Versioning-based Isolation Levels