The locks that SQL Anywhere uses when a user enters a SELECT statement depend on the transaction's isolation level. All SELECT statements, regardless of isolation level, acquire schema locks on the referenced tables.
No locking operations are required when executing a SELECT statement at isolation level 0. Each transaction is not protected from changes introduced by other transactions. It is the responsibility of the programmer or database user to interpret the result of these queries with this limitation in mind.
SQL Anywhere does not use many more locks when running a transaction at isolation level 1 than it does at isolation level 0. The database server modifies its operation in only two ways.
The first difference in operation has nothing to do with acquiring locks, but rather with respecting them. At isolation level 0, a transaction can read any row, even if another transaction has acquired a write lock. By contrast, before reading each row, an isolation level 1 transaction must check whether a write lock is in place. It cannot read past any write-locked rows because doing so might entail reading dirty data. The use of the READPAST hint permits the server to ignore write-locked rows, but while the transaction will no longer block, its semantics no longer coincide with those of isolation level 1. See READPAST hint in FROM clause.
The second difference in operation affects cursor stability. Cursor stability is achieved by acquiring a short-term read lock on the current row of a cursor. This read lock is released when the cursor is moved. More than one row may be affected if the contents of the cursor is the result of a join. In this case, the database server acquires short-term read locks on all rows which have contributed information to the cursor's current row, and releases these locks when another row of the cursor is selected as the current row.
At isolation level 2, the database server modifies its operation to ensure repeatable read semantics. If a SELECT statement returns values from every row in a table, then the database server acquires a read lock on each row of the table as it reads it. If, instead, the SELECT contains a WHERE clause, or another condition which restricts the rows in the result, then the database server instead reads each row, tests the values in the row against that condition, and then acquires a read lock on the row if it meets that condition. The read locks that are acquired are long-term read locks and are held until the transaction completes via an implicit or explicit COMMIT or ROLLBACK statement. As with isolation level 1, cursor stability is assured at isolation level 2, and dirty reads are not permitted.
When operating at isolation level 3, the database server is obligated to ensure that all transaction schedules are serializable. In particular, in addition to the requirements imposed at isolation level 2, it must prevent phantom rows so that re-executing the same statement is guaranteed to return the same results in all circumstances.
To accommodate this requirement, the database server uses read locks and phantom locks. When executing a SELECT statement at isolation level 3, the database server acquires a read lock on each row that is processed during the computation of the result set. Doing so ensures that no other transactions can modify those rows until the transaction completes.
This requirement is similar to the operations that the database server performs at isolation level 2, but differs in that a lock must be acquired for each row read, whether those rows satisfy any predicates in the SELECT's WHERE, ON, or HAVING clauses. For example, if you select the names of all employees in the sales department, then the server must lock all the rows which contain information about a sales person, whether the transaction is executing at isolation level 2 or 3. At isolation level 3, however, the server must also acquire read locks on each of the rows of employees which are not in the sales department. Otherwise, another transaction could potentially transfer another employee to the sales department while the first transaction was still executing.
There are two implications when a read lock must be acquired for each row read:
The database server may need to place many more locks than would be necessary at isolation level 2. The number of phantom locks acquired is one more than the number of read locks that are acquired for the scan. This doubling of the lock overhead adds to the execution time of the request.
The acquisition of read locks on each row read has a negative impact on the concurrency of database update operations to the same table.
The number of phantom locks the database server acquires can vary greatly and depends upon the execution strategy chosen by the query optimizer. The SQL Anywhere query optimizer will attempt to avoid sequential scans at isolation level 3 because of the potentially adverse affects on overall system concurrency, but the optimizer's ability to do so depends upon the predicates in the statement and on the relevant indexes available on the referenced tables.
As an example, suppose you want to select information about the employee with Employee ID 123. As EmployeeID is the primary key of the employee table, the query optimizer will almost certainly choose an indexed strategy, using the primary key index, to locate the row efficiently. In addition, there is no danger that another transaction could change another Employee's ID to 123 because primary key values must be unique. The server can guarantee that no second employee is assigned that ID number simply by acquiring a read lock on the row containing information about employee 123.
In contrast, the database server would acquire more locks were you instead to select all the employees in the sales department. In the absence of a relevant index, the database server must read every row in the employee table and test whether each employee is in sales. If this is the case, both read and phantom locks must be acquired for each row in the table.
SELECT statements that execute at snapshot, statement-snapshot, or readonly-statement-snapshot do not acquire read locks. This is because each snapshot transaction (or statement) sees a snapshot of a committed state of the database at some previous point in time. The specific point in time is determined by which of the three snapshot isolation levels is being used by the statement. As such, read transactions never block update transactions and update transactions never block readers. Therefore, snapshot isolation can give considerable concurrency benefits in addition to the obvious consistency benefits. However, there is a tradeoff; snapshot isolation can be very expensive. This is because the consistency guarantee of snapshot isolation means that copies of changed rows must be saved, tracked, and (eventually) deleted for other concurrent transactions.
|Discuss this page in DocCommentXchange.
Send feedback about this page using email.
|Copyright © 2009, iAnywhere Solutions, Inc. - SQL Anywhere 11.0.1|