Chapter-3

Transaction Management and

Concurrency Control

1
What is a Transaction?
 Any action that reads from and/or writes to a
database. It may consist of:
 Simple SELECT statement to generate list of
table contents
 Series of related UPDATE statements to change
values of attributes in various tables
 Series of INSERT statements to add rows to one
or more tables
 Combination of SELECT, UPDATE, and INSERT
statements 2
What is a Transaction? (continued)

 Transaction is logical unit of work that must be

either entirely completed or aborted
 Successful transactions change the database from
one consistent state to another
 One in which all data integrity constraints are
satisfied
 Most real-world database transactions are formed
by two or more database requests
 Equivalent of a single SQL statement in an
application program or transaction

3
Evaluating Transaction Results

 Not all transactions update database

 SQL code represents a transaction because
database was accessed
 Improper or incomplete transactions can have
devastating effect on database integrity
 Some DBMSs provide means by which user
can define enforceable constraints
 Other integrity rules are enforced
automatically by the DBMS

4
Transaction Properties (ACIDS)

 Atomicity
 Requires that all operations (SQL requests) of
a transaction be completed or none are
completed.
 Consistency
 The database must begin in a consistent state
and end in a consistent state
 Consistent state = all integrity constraints satisfied.

5
Transaction Properties (continued)
 Isolation
 Data used during execution of a transaction
cannot be used by second transaction until
first one is completed
 Durability
 Transactionchanges that are saved to the
database cannot be lost even if the system
crashes.
 No changes are saved to the database until the
transaction completes successfully.

6
Transaction Properties (continued)

 Serializability
 Ensures that concurrent execution of several
transactions yields consistent results

7
Transaction Management with SQL

 ANSI has defined standards that govern

SQL database transactions
 Transaction support is provided by two
SQL statements: COMMIT and
ROLLBACK

8
Transaction Management with SQL
(continued)
 ANSI standards require that, when a
transaction sequence is initiated by a user
or an application program, it must continue
through all succeeding SQL statements
until one of four events occurs
 COMMIT statement is reached
 ROLLBACK statement is reached
 End of program is reached
 Program is abnormally terminated
9
The Transaction Log
 A DBMS uses a transaction log to keep track
of all transactions that update the database.
 Transaction log stores:
A record for the beginning of transaction
 For each transaction component (SQL statement):
 Type of operation being performed (update, delete,
insert)
 Names of objects affected by transaction
 “Before” and “after” values for updated fields
 Pointers to previous and next transaction log entries for
the same transaction
 Ending (COMMIT) of the transaction

10
The Transaction Log (continued)

11
Concurrency Control
 Coordination of simultaneous transaction
execution in a multiprocessing database
system
 Objective is to ensure serializability of
transactions in a multiuser database
environment

12
Concurrency Control (continued)

 Simultaneous execution of transactions

over a shared database can create several
data integrity and consistency problems
 Lostupdates
 Uncommitted data
 Inconsistent retrievals

13
Lost Updates

14
Lost Updates (continued)

15
Uncommitted Data

16
Uncommitted Data (continued)

17
Inconsistent Retrievals

18
Inconsistent Retrievals
(continued)

19
Inconsistent Retrievals
(continued)

20
The Scheduler
 Special DBMS program
 Purpose is to establish order of operations
within which concurrent transactions are
executed
 Interleaves execution of database
operations to ensure serializability and
isolation of transactions

21
The Scheduler (continued)
 Bases its actions on concurrency control
algorithms
 Ensures computer’s central processing
unit (CPU) is used efficiently
 Facilitates data isolation to ensure that two
transactions do not update the same data
element at the same time

22
The Scheduler (continued)

23
The Scheduler (continued)
 The scheduler bases its actions on
concurrency control algorithms, such
as locking or time stamping methods to
determine the appropriate order

24
Concurrency Control with Locking Methods

 Lock
 Guarantees exclusive use of a data item to a
current transaction
 Required to prevent another transaction from
reading inconsistent data
 Lock manager
 Responsiblefor assigning and policing the
locks used by transactions

25
Lock Granularity
 Indicates level of lock use
 Locking can take place at following levels:
 Database
 Table
 Page
 Row
 Field (attribute)

26
Lock Granularity (continued)
 Database-level lock
 Entire database is locked
 Table-level lock
 Entire table is locked
 Page-level lock
 Entire diskpage is locked

27
Lock Granularity (continued)
 A diskpage (page) is the equivalent of a
diskblock which is directly addressable section
of a disk.
 A page has a fixed size, such as 4k, 8k, or 16k.
 E.g. if you want to write only 73 bytes to a 4k
page, the entire 4k must be read from disk,
updated in memory and written back to disk
 Page-level locks are currently the most used
multi-user DBMS locking methods.

28
Lock Granularity (continued)
 Row-level lock
 Allows concurrent transactions to access different
rows of same table, even if rows are located on same
page
 A row-level lock is less restrictive than the
previous locking methods.
 Advantage: it improves the availability of data.
 Disadvantage: its management require high overhead
(a lock exists for every row in a table).
 Field-level lock
 Allows concurrent transactions to access same row,
as long as they require use of different fields
(attributes) within that row

29
Lock Types
 Binary lock
 Has only two states: locked (1) or unlocked (0)
 Exclusive lock
 Access is specifically reserved for transaction that
locked object
 Must be used when potential for conflict exists
 Issued when a transaction needs to update data and
no locks are currently held on it.
 Shared lock
 Concurrent transactions are granted Read access on
basis of a common lock
30
Lock Types (continued)
 Although locks prevent serious data
inconsistencies, they can lead to two major
problems:
 The resulting transaction schedule may not be
serializable.
 The schedule may create deadlocks.

31
Two-Phase Locking to Ensure
Serializability
 Defines how transactions acquire and relinquish
locks
 Guarantees serializability, but it does not prevent
deadlocks
 Growing phase - Transaction acquires all
required locks without unlocking any data
 Shrinking phase - Transaction releases all
locks and cannot obtain any new lock

32
Two-Phase Locking to Ensure Serializability
(continued)

 The two-phase locking protocol is governed by

the following rules:
 Two transactions cannot have conflicting locks
 No unlock operations can precede a lock operation in
the same transaction.
 No data are affected until all locks are obtained – that
is until the transaction is in its locked point.

33
Two-Phase Locking to Ensure Serializability
(continued)

34
Deadlocks
 Condition that occurs when two transactions
wait for each other to unlock data
 Possible only if one of the transactions wants
to obtain an exclusive lock on a data item
 No deadlock condition can exist among
shared locks

35
Deadlocks (continued)

36
Deadlocks (continued)

 Controlled through:
 Deadlock Prevention
 Deadlock Detection
 Deadlock Avoidance

37
Concurrency Control with Time Stamping
Methods
 Assigns global unique time stamp to each transaction
 Produces explicit order in which transactions are
submitted to the DBMS
 Uniqueness
 Ensures that no equal time stamp values can exist
 Monotonicity
 Ensures that time stamp values always increase

 All database operations (read and write) within

the same transaction must have the same time
stamp.
38
Concurrency Control with Optimistic Methods

 Optimistic approach
 Based on assumption that majority of
database operations do not conflict
 Does not require locking or time stamping
techniques
 Transaction is executed without restrictions
until it is committed
 Phases are read, validation, and write

39