Chapter 9

Transaction Management and

Concurrency Control
What is a Transaction?
 Any action that reads from and/or writes
to a database may consist of
 Simple SELECT statement to generate a list of
table contents
 A series of related UPDATE statements to
change the values of attributes in various
tables
 A series of INSERT statements to add rows to
one or more tables
 A combination of SELECT, UPDATE, and
INSERT statements

Database Systems 6e/ Rob & Coronel Slide 2

What is a Transaction?

 A logical unit of work that must be either

entirely completed or aborted
 Successful transaction changes 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
 The equivalent of a single SQL statement in an
application program or transaction

Database Systems 6e/ Rob & Coronel Slide 3

The Relational Schema
for the Ch09_SaleCo Database

Database Systems 6e/ Rob & Coronel Slide 4

 Evaluating Transaction Results
 Not all transactions update the database
 SQL code represents a transaction because
database was accessed
 Improper or incomplete transactions can have
a devastating effect on database integrity
 Some DBMSs provide means by which user can
define enforceable constraints based on
business rules
 Other integrity rules are enforced automatically
by the DBMS when table structures are properly
defined, thereby letting the DBMS validate some
transactions

Database Systems 6e/ Rob & Coronel Slide 5

Transaction Processing

 For example, a transaction may involve

 The creation of a new invoice
 Insertion of an row in the LINE table
 Decreasing the quantity on hand by 1
 Updating the customer balance
 Creating a new account transaction row
 If the system fails between the first and
last step, the database will no longer be
in a consistent state Figure 9.2

Database Systems 6e/ Rob & Coronel Slide 6

Transaction Properties

 Atomicity
 Requires that all operations (SQL requests) of
a transaction be completed; if not, then the
transaction is aborted
 A transaction is treated as a single, indivisible,
logical unit of work
 Durability
 Indicates permanence of database’s consistent
state
 When a transaction is complete, the database
reaches a consistent state. That state can not
be lost even if the system fails

Database Systems 6e/ Rob & Coronel Slide 7

 Transaction Properties
 Serializability
 Ensures that the concurrent execution of
several transactions yields consistent results
 Multiple, concurrent transactions appear as if
they executed in serial order (one after
another)
 Isolation
 Data used during execution of a transaction
cannot be used by second transaction until first
one is completed

Database Systems 6e/ Rob & Coronel Slide 8

Transaction Properties
 Consistency
 Indicates the permanence of database’s
consistent state

Database Systems 6e/ Rob & Coronel Slide 9

Transaction Management with SQL

 ANSI has defined standards that govern

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

Database Systems 6e/ Rob & Coronel Slide 10

Transaction Management with SQL

1. A COMMIT statement is reached- all changes

are permanently recorded within the database
2. A ROLLBACK is reached – all changes are
aborted and the database is restored to a
previous consistent state
3. The end of the program is successfully
reached – equivalent to a COMMIT
4. The program abnormally terminates and a
rollback occurs

Database Systems 6e/ Rob & Coronel Slide 11

Transaction States
 Active – the initial state; the transaction stays in this state
while it is executing
 Partially committed – after the final statement has been
executed.
 Failed – after the discovery that normal execution can no
longer proceed.
 Aborted – after the transaction has been rolled back and the
database restored to its state prior to the start of the
transaction. Two options after it has been aborted:
 restart the transaction
 can be done only if no internal logical error
 kill the transaction
 Committed – after successful completion.

 To guarantee atomicity, external observable action should all

be performed (in order) after the transaction is committed.

Database Systems 6e/ Rob & Coronel Slide 12

States Transition Diagram

Partially
committed committed

active

failed aborted

Database Systems 6e/ Rob & Coronel Slide 13

 The Transaction Log
 Keeps track of all transactions that updatethe
database. It contains:
 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 the transaction (the name
of the table)
 “Before” and “after” values for updated fields
 Pointers to previous and next transaction log entries for
the same transaction
 The ending (COMMIT) of the transaction
 Increases processing overhead but the ability to restore a
corrupted database is worth the price

Database Systems 6e/ Rob & Coronel Slide 14

 The Transaction Log
 Increases processing overhead but the ability to
restore a corrupted database is worth the price
 If a system failure occurs, the DBMS will examine
the log for all uncommitted or incomplete
transactions and it will restore the database to a
previous state
 The log it itself a database and to maintain its
integrity many DBMSs will implement it on several
different disks to reduce the risk of system failure

Database Systems 6e/ Rob & Coronel Slide 15

A Transaction Log

Database Systems 6e/ Rob & Coronel Slide 16

 Concurrency Control
 The coordination of the simultaneous execution of
transactions in a multiprocessing database is
known as concurrency control

 The objective of concurrency control is to ensure

the serializability of transactions in a multiuser
database environment

Database Systems 6e/ Rob & Coronel Slide 17

Concurrency Control

 Important  simultaneous execution of

transactions over a shared database can
create several data integrity and consistency
problems
 The three main problems are:

 lost updates

 uncommitted data

 inconsistent retrievals

Database Systems 6e/ Rob & Coronel Slide 18

Normal Execution of Two Transactions

Database Systems 6e/ Rob & Coronel Slide 19

Lost Updates

Database Systems 6e/ Rob & Coronel Slide 20

Uncommited Data Problem

 Uncommitted data occurs when two

transactions execute concurrently and
the first is rolled back after the second
has already accessed the uncommitted
data
 This violates the isolation property of
transactions

Database Systems 6e/ Rob & Coronel Slide 21

Correct Execution of Two Transactions

Database Systems 6e/ Rob & Coronel Slide 22

An Uncommitted Data Problem

Database Systems 6e/ Rob & Coronel Slide 23

Inconsistent Retrieval Problem

 Occur when a transaction calculates some

aggregate functions over a set of data while
transactions are updating the data
 Some data may be read after they are changed and
some before they are changed yielding inconsistent
results

Database Systems 6e/ Rob & Coronel Slide 24

Retrieval During Update

Database Systems 6e/ Rob & Coronel Slide 25

Transaction Results: Data Entry Correction

Database Systems 6e/ Rob & Coronel Slide 26

Inconsistent Retrievals

Database Systems 6e/ Rob & Coronel Slide 27

The Scheduler

 Special DBMS program: establishes

order of operations within which
concurrent transactions are executed
 Interleaves the execution of database
operations to ensure serializability and
isolation of transactions
 To determine the appropriate order, the
scheduler bases its actions on
concurrency control algorithms such as
locking and time stamping

Database Systems 6e/ Rob & Coronel Slide 28

The Scheduler (continued)

 Ensures computer’s central processing

unit (CPU) is used efficiently
 Default would be FIFO without preemption
– idle CPU (during I/O) is inefficient use of
the resource and result in unacceptable
response times in within the multiuser
DBMS environment
 Facilitates data isolation to ensure that
two transactions do not update the
same data element at the same time

Database Systems 6e/ Rob & Coronel Slide 29

Read/Write Conflict Scenarios:
Conflicting Database Operations Matrix

 Transactions T1 and T2 are executing concurrently over the

same data
 When they both access the data and at least one is executing a
WRITE, a conflict can occur

Database Systems 6e/ Rob & Coronel Slide 30

Concurrency Control with Locking Methods
 Lock
 Guarantees exclusive use of a data item to a
current transaction
 T2 does not have access to a data item that is
currently being used by T1
 Transaction acquires a lock prior to data access; the
lock is released when the transaction is complete
 Required to prevent another transaction from
reading inconsistent data
 Lock manager
 Responsible for assigning and policing the locks
used by the transactions

Database Systems 6e/ Rob & Coronel Slide 31

Lock Granularity
 Indicates the level of lock use
 Locking can take place at the following levels:
 Database-level lock
 Entire database is locked
 Table-level lock
 Entire table is locked
 Page-level lock
 Entire diskpage is locked
 Row-level lock
 Allows concurrent transactions to access different rows
of the same table, even if the rows are located on the
same page
 Field-level lock
 Allows concurrent transactions to access the same row,
as long as they require the use of different fields
(attributes) within that row

Database Systems 6e/ Rob & Coronel Slide 32

A Database-Level Locking Sequence

 Good for batch processing but unsuitable for online multi-user

DBMSs
 T1 and T2 can not access the same database concurrently even if
they use different tables

Database Systems 6e/ Rob & Coronel Slide 33

Table-Level Lock

 T1 and T2 can access the same database concurrently as long as they use different tables
 Can cause bottlenecks when many transactions are trying to access the same table (even if the
transactions want to access different parts of the table and would not interfere with each other)
 Not suitable for multi-user DBMSs

Database Systems 6e/ Rob & Coronel Slide 34

Page-Level Lock

 An entire disk page is locked (a table can span several pages and
each page can contain several rows of one or more tables)
 Most frequently used multi-user DBMS locking method

Database Systems 6e/ Rob & Coronel Slide 35

Row-Level Lock

 Concurrent transactions can access different rows of the same

table even if the rows are located on the same page
 Improves data availability but with high overhead (each row has
a lock that must be read and written to)

Database Systems 6e/ Rob & Coronel Slide 36

Field-Level Lock
 Allows concurrent transactions to
access the same row as long as they
require the use of different fields within
that row
 Most flexible lock but requires an
extremely high level of overhead

Database Systems 6e/ Rob & Coronel Slide 37

Binary Locks
 Has only two states: locked (1) or unlocked (0)
 Eliminates “Lost Update” problem – the lock is not released until
the write statement is completed
 Can not use PROD_QOH until it has been properly updated
 Considered too restrictive to yield optimal concurrency conditions as it
locks even for two READs when no update is being done

Database Systems 6e/ Rob & Coronel Slide 38

Shared/Exclusive Locks
 Exclusive lock
 Access is specifically reserved for the transaction that locked the
object
 Used when potential for conflict exists – when a transaction wants to
update a data item and no other transaction currently hold locks on
that data item
 Granted if and only if no other locks are held on the data item
 Shared lock
 Concurrent transactions are granted Read access on the basis of a
common lock
 Issued when a transaction wants to read data and no exclusive lock is
held on that data item
 Multiple transactions can each have a shared lock on the same data
item if they are all just reading it
 Mutual Exclusive Rule
 Only one transaction at a time can own an exclusive lock on the same
object

Database Systems 6e/ Rob & Coronel Slide 39

Shared/Exclusive Locks
 Increased overhead
 The type of lock held must be known before a lock can be
granted
 Three lock operations exist:
 READ_LOCK to check the type of lock
 WRITE_LOCK to issue the lock
 UNLOCK to release the lock
 A lock can be upgraded from share to exclusive and
downgraded from exclusive to share
 Two possible major problems may occur
 The resulting transaction schedule may not be serializable
 The schedule may create deadlocks

Database Systems 6e/ Rob & Coronel Slide 40

Two-Phase Locking Protocol for Serializability
 Defines how transactions acquire and
relinquish locks
 Guarantees serializability, but it does not
prevent deadlocks
 Growing phase, in which a transaction acquires
all the required locks without unlocking any
data
 Shrinking phase, in which a transaction
releases all locks and cannot obtain any new
lock

Database Systems 6e/ Rob & Coronel Slide 41

Two-Phase Locking to Ensure Serializability
 Governed by the following rules:
 Two transactions cannot have conflicting locks

 No unlock operation 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

Database Systems 6e/ Rob & Coronel Slide 42

Two-Phase Locking Protocol

Database Systems 6e/ Rob & Coronel Slide 43

Deadlocks
 Condition that occurs when two transactions wait for
each other to unlock data
 T1 needs data items X and Y; T needs Y and X
 Each gets the its first piece of data but then waits to get the
second (which the other transaction is already holding) –
deadly embrace
 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
 Control through
 Prevention
 Detection
 Avoidance

Database Systems 6e/ Rob & Coronel Slide 44

How a Deadlock Condition Is Created

Database Systems 6e/ Rob & Coronel Slide 45

Deadlock Detection

Wait-for graph without a cycle Wait-for graph with a cycle

 Deadlocks can be described as a wait-for graph where:

 vertices are all the transactions in the system
 There is an edge Ti Tk in case Ti is waiting for Tk
 When Ti requests a data item currently being held by Tk, then the
edge Ti Tk is inserted in the wait-for graph. This edge is removed
only when Tk is no longer holding a data item needed by Ti.
 The system is in a deadlock state if and only if the wait-for graph
has a cycle. Must invoke a deadlock-detection algorithm
periodically to look for cycles.

Database Systems 6e/ Rob & Coronel Slide 46

Pitfalls of Lock-Based Protocols
 Consider the partial schedule

 Neither T3 nor T4 can make progress — executing lock-S(B)

causes T4 to wait for T3 to release its lock on B, while executing
lock-X(A) causes T3 to wait for T4 to release its lock on A.
 Such a situation is called a deadlock.
 To handle a deadlock one of T3 or T4 must be rolled back
and its locks released.

Database Systems 6e/ Rob & Coronel Slide 47

 Pitfalls of Lock-Based Protocols (Cont.)

 The potential for deadlock exists in most

locking protocols. Deadlocks are a necessary
evil.
 Starvation is also possible if concurrency
control manager is badly designed. For
example:
 A transaction may be waiting for an X-lock on an
item, while a sequence of other transactions
request and are granted an S-lock on the same
item.
 The same transaction is repeatedly rolled back
due to deadlocks.

Database Systems 6e/ Rob & Coronel Slide 48

 Pitfalls of Lock-Based Protocols (Cont.)

 Concurrency control manager can be

designed to prevent starvation.
 Two-phase locking does not ensure freedom
from deadlocks
 Deadlock prevention protocols or deadlock
detection mechanisms are needed!
 With detection mechanisms when deadlock is
detected :
 Some transaction will have to rolled back (made
a victim) to break deadlock. Select that
transaction as victim that will incur minimum
cost.

Database Systems 6e/ Rob & Coronel Slide 49

Concurrency Control with Time
Stamping Methods
 Assigns a global unique time stamp to each transaction
 All database operations within the same transaction must have
the same time stamp
 Produces an 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
 Disadvantage
 Each value stored in the database requires two additional time
stamp fields – last read, last update

Database Systems 6e/ Rob & Coronel Slide 50

Wait/Die and Wound/Wait Schemes
 Wait/die
 Older transaction waits and the younger is rolled back
and rescheduled

 Wound/wait
 Older transaction rolls back the younger transaction
and reschedules it

 In the situation where a transaction requests

multiple locks, each lock request has an
associated time-out value. If the lock is not
granted before the time-out expires, the
transaction is rolled back

Database Systems 6e/ Rob & Coronel Slide 51

Wait/Die and Wound/Wait
Concurrency Control Schemes

Database Systems 6e/ Rob & Coronel Slide 52

Concurrency Control with Optimistic Methods
 Optimistic approach
 Based on the assumption that the majority of
database operations do not conflict
 Does not require locking or time stamping
techniques
 Transaction is executed without restrictions until it
is committed
 Acceptable for mostly read or query database
systems that require very few update transactions
 Phases are read, validation, and write

Database Systems 6e/ Rob & Coronel Slide 53

Concurrency Control with Optimistic Methods
 Phases are read, validation, and write
 Read phase – transaction reads the database, executes
the needed computations and makes the updates to a
private copy of the database values.
 All update operations of the transaction are recorded in
a temporary update file which is not accessed by the
remaining transactions
 Validation phase – transaction is validated to ensure
that the changes made will not affect the integrity and
consistency of the database
 If the validation test is positive, the transaction goes to
the writing phase. If negative, the transaction is
restarted and the changes discarded
 Writing phase – the changes are permanently applied
to the database

Database Systems 6e/ Rob & Coronel Slide 54

Database Recovery

Backup Levels:
Full Backup – save whole
database plus transaction log
Differential Backup (save
only pages that changed since last backup)
Transaction Log Backup
(save only the transaction log)

Database Systems 6e/ Rob & Coronel Slide 55

Database Recovery
 Database / System Failures:
 Software (O.S., DBMS, Application
Programs, Viruses)
 Hardware (Memory Chips, Disk
Crashes, Bad Sectors)
 Programming Exemption (Application
Program rollbacks)
 Transaction (Aborting transactions
due to deadlock detection)
 External (Fire, Flood, sabotage, etc.)

Database Systems 6e/ Rob & Coronel Slide 56

Transaction Recovery
 The database recovery process involves
bringing the database to a consistent state
after failure.
 Transaction recovery procedures generally
make use of deferred-write and write-
through techniques

Database Systems 6e/ Rob & Coronel Slide 57

Transaction Recovery
 Write-through
 Database is immediately updated by
transaction operations during the transaction’s
execution, even before the transaction reaches
its commit point
 If the transaction aborts before it reaches its
commit point, a ROLLBACK is done to restore
the database to a consistent state
 A transaction that committed after the last checkpoint is
redone using the “after” values of the log
 A transaction with a ROLLBACK after the last checkpoint
is rolled back using the “before” values in the log

Database Systems 6e/ Rob & Coronel Slide 58

Transaction Recovery
 Deferred write
 Transaction operations do not
immediately update the physical
database
 Only the transaction log is updated
 When updated buffers are being
written to disk, all other requests are
halted, not executed

Database Systems 6e/ Rob & Coronel Slide 59

Transaction Recovery
 Deferred write, continued…
 Redundant Transaction Logs – keep
several copies of log on different
devices, to offset failure.
 Database Buffers – Buffers are used
to hold transaction logs and
dirty/modified pages
 they increase processing time (overhead)
on updates especially when redundancy
is involved.

Database Systems 6e/ Rob & Coronel Slide 60

Transaction Recovery
 Recover Database by using data
in the Transaction Log
 If the transaction aborts before it
reaches its commit point, no
ROLLBACK is needed because the DB
was never updated
 A transaction that performed a
COMMIT after the last checkpoint is
redone using the “after” values of the
transaction log
Database Systems 6e/ Rob & Coronel Slide 61
Database Recovery Management
 Database recovery
 Restores database from a given state, usually
inconsistent, to a previously consistent state
 Based on the atomic transaction property
 All portions of the transaction must be treated as
a single logical unit of work, in which all
operations must be applied and completed to
produce a consistent database
 If transaction operation cannot be completed,
transaction must be aborted, and any changes
to the database must be rolled back (undone)

Database Systems 6e/ Rob & Coronel Slide 62

Checkpoints
 Checkpoints are suited for deferred write strategy.
 They are records registered in the transaction log.
 They mark the last consistent state of the database.
 They help reduce the time taken to recover from a
crash.
 only transactions that executed since last checkpoint are
reviewed.
 Writing updated buffers to the disk is scheduled
several times per hour
 Saves information about active transactions and dirty
buffer pool pages

Database Systems 6e/ Rob & Coronel Slide 63

A Transaction Log for Transaction
Recovery Examples

Database Systems 6e/ Rob & Coronel Slide 64

Transaction Recovery Examples
 T101 consists of 2 UPDATE statements that reduce
the QOH for product 54778-2T and increase the
customer balance for customer 10011 for a credit
sale of 2 units of that product
 T106 represents a credit sale of 1 unit of 89-WRE-
Q to customer 10016 for $277.55 This transaction
consists of 3 INSERT and 2 UPDATE statements
 T155 is an inventory update increasing QOH for
2232/QWE to 26 units
 A DB checkpoint wrote all updated database
buffers to disk which is only for T101

Database Systems 6e/ Rob & Coronel Slide 65

Transaction Recovery Examples
 Last checkpoint was 423
 T101 started and finished before that checkpoint
so all changes were written to disk and no action
need be taken
 T106 and T155 committed after the last
checkpoint so they will have their “after” values
written to disk using the log

Database Systems 6e/ Rob & Coronel Slide 66

ARIES Recovery - Algorithms for Recovery and
Isolation Exploiting Semantics

 There are 3 phases in the Aries recovery

algorithm:
 Analysis: Scan the log forward (from
the most recent checkpoint) to identify
all transactions that were active, and all
dirty pages in the buffer pool at the
time of the crash.
 Redo: Redo all updates to dirty pages
in the buffer pool, as needed, to ensure
that all logged updates are in fact
carried out and written to disk.
Database Systems 6e/ Rob & Coronel Slide 67
ARIES Recovery - Algorithms for Recovery and
Isolation Exploiting Semantics

 There are 3 phases in the Aries recovery

algorithm…continued:
 Undo: The writes of all transactions
that were active at the crash are
undone
 By restoring the before value of the update,
which is in the log record for the update.
 working backwards in the log.
 Some care must be taken to handle the
case of a crash occurring during the
recovery process!

Database Systems 6e/ Rob & Coronel Slide 68