TRANSACTION MANAGEMENT

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.1

 Chapters 15-17: Transaction Management
 Transaction (Chapter 15)
 Transaction Concept
 Transaction State
 Concurrent Executions
 Serializability
 Recoverability
 Testing for Serializability
 Concurrency control (Chapter 16)
 Lock-based protocols
 Timestamp-based protocols
 Multiple granularity
 Multiversion schemes
 Recovery Systems (Chapter 17)
 Log-based recovery
 Recovery with concurrent transactions
 Transaction in SQL
 Transaction management in Oracle 10g

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.2

 Transaction Concept
 A transaction is a unit of program execution that accesses and possibly
updates various data items.
 A transaction is a program including a collection of database operations,
executed as a logical unit of data processing.
 Each high level operation can be divided into a number of low level tasks
or operations. For example, a data update operation can be divided into
three tasks −
 read_item() − reads data item from storage to main memory.
 modify_item() − change value of item in the main memory.
 write_item() − write the modified value from main memory to storage
 E.g. transaction to transfer €50 from account A to account B:
1. read_from_account(A)
2. A := A – 50
3. write_to_account(A) Transaction
4. read_from_account(B)
5. B := B + 50
6. write_to_account(B)
José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.3
 • Two main issues to deal with:
Failures of various kinds, such as hardware failures and
system crashes
Concurrent execution of multiple transactions

Transaction Operations
 The low level operations performed in a transaction are −
 begin_transaction − A marker that specifies start of
transaction execution.
 read_item or write_item − Database operations that may
be interleaved with main memory operations as a part of
transaction.
 end_transaction − A marker that specifies end of
transaction.
 commit − A signal to specify that the transaction has been
successfully completed in its entirety and will not be
undone.
 rollback − A signal to specify that the transaction has been
unsuccessful and so all temporary changes in the database
are
José Alferes - Adaptado de undone.
Database A committed
System Concepts
th
- 5 Edition transaction
15-17.4 cannot be rolled back.
 Transaction ACID properties
 E.g. transaction to transfer €50 from account A to account B:
1. read_from_acoount(A) (Suppose A=100,B=200 A+B=300)
2. A := A – 50 (A=50)
3. write_to_account(A) <---(Power Failure)
4. read_from_accont(B)
Mismatch i.e.
5. B := B + 50 data
6. write_to_account(B) (A+B=50+200=250) inconsistent
 Atomicity requirement
 if the transaction fails after step 3 and before step 6, money will be “lost”
leading to an inconsistent database state
 Failure could be due to software or hardware
 the system should ensure that updates of a partially executed transaction
are not reflected in the database
 All or nothing, regarding the execution of the transaction
 Durability requirement — once the user has been notified of transaction has
completion, the updates must persist in the database even if there are software
or hardware failures.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.5

 Transaction ACID properties (Cont.)
 Transaction to transfer €50 from account A to account B:
1. read_from_acoount(A) (Suppose A=100,B=200 A+B=300)
2. A := A – 50 (A=100-50=50)
3. write_to_account(A)
4. read_from_accont(B) Match i.e. data
5. B := B + 50 (B=200+50=250) consistent
6. write_to_account(B) (A+B=50+250=300)
 Consistency requirement in above example:
 the sum of A and B is unchanged by the execution of the transaction
 In general, consistency requirements include
 Explicitly specified integrity constraints such as primary keys and foreign
keys
 Implicit integrity constraints
– e.g. sum of balances of all accounts, minus sum of loan amounts
must equal value of cash-in-hand
 A transaction must see a consistent database and must leave a consistent
database
 During transaction execution the database may be temporarily inconsistent.
 Constraints to be verified only at the end of the transaction

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.6

 Transaction ACID properties (Cont.)
 Isolation requirement — if between steps 3 and 6, another
transaction T2 is allowed to access the partially updated database, it
will see an inconsistent database (the sum A + B will be less than it
should be).
T1 T2
1. read(A) (A=100)
2. A := A – 50 (A=50)
3. write(A)
read(A), read(B), print(A+B) (A+B=50+200=250)
4. read(B)
5. B := B + 50
6. write(B)
 Isolation can be ensured trivially by running transactions serially
 that is, one after the other.
 However, executing multiple transactions concurrently has significant
benefits, as we will see later.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.7

 ACID Properties - Summary
A transaction is a unit of program execution that accesses and
possibly updates various data items.To preserve the integrity of data
the database system must ensure:
 Atomicity Either all operations of the transaction are properly reflected
in the database or none are.
 Consistency Execution of a (single) transaction preserves the
consistency of the database.
 Isolation Although multiple transactions may execute concurrently,
each transaction must be unaware of other concurrently executing
transactions. Intermediate transaction results must be hidden from other
concurrently executed transactions.
 That is, for every pair of transactions Ti and Tj, it appears to Ti that
either Tj, finished execution before Ti started, or Tj started execution
after Ti finished.
 Durability. After a transaction completes successfully, the changes it
has made to the database persist, even if there are system failures.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.8

 Transaction State
 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.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.9

 Transaction State (Cont.)

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.10

 Concurrent Executions
 Multiple transactions are allowed to run concurrently in the system.
Advantages are:
 increased processor and disk utilization, leading to better
transaction throughput
 E.g. one transaction can be using the CPU while another is
reading from or writing to the disk
 reduced average response time for transactions: short
transactions need not wait behind long ones.
 Concurrency control schemes – mechanisms to achieve isolation
 that is, to control the interaction among the concurrent
transactions in order to prevent them from destroying the
consistency of the database
 Two-phase lock protocol
 Timestamp-Based Protocols
 Validation-Based Protocols
 Studied in Operating Systems, and briefly summarized later

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.11

 Schedules
 Schedule – a sequences of instructions that specify the chronological
order in which instructions of concurrent transactions are executed
 a schedule for a set of transactions must consist of all instructions
of those transactions
 must preserve the order in which the instructions appear in each
individual transaction.
 A transaction that successfully completes its execution will have a
commit instructions as the last statement
 by default transaction assumed to execute commit instruction as its
last step
 A transaction that fails to successfully complete its execution will have
an abort instruction as the last statement

 The goal is to find schedules that preserve the consistency.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.12

 Example Schedule 1
 Let T transfer €50 from A to B, and T transfer 10% of the
1 2
balance from A to B.
 A serial schedule in which T is followed by T :
1 2

A=100
A=50

B=200
B=250
SUM=300
A= 50
temp=50*.10=5
A=45
B=250
B=250+5=255

A+B=300
José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.13
 Equivalent Schedule

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.14

 Example Schedule 2
 A serial schedule where T2 is followed by T1

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.15

 Example Schedule 3
 Let T1 and T2 be the transactions defined previously. The
following schedule is not a serial schedule, but it is equivalent
to Schedule 1.
SupposeA=100,B=200
A=100
A=50

temp=50*0.1=5
A=50-5=45

B=200
B=200+50=250

B=250
B=250+5=255
A+B=45+255=300

In Schedules 1, 2 and 3, the sum A + B is preserved.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.16
 Example Schedule 4
 The following concurrent schedule does not preserve the
value of (A + B ).
SupposeA=100,B=200
A=100
A=50
A=100
temp=100*.10=10
A=100-10=90
B=200

A=90
B=200
B=200+50
B=250
B=250+10=260

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.17

 Serializability

T1 T2
R(A)
W(A)

R(A)
W(A)

T1
T2 Precedence Graph

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.18

 Serializability
T1 T2
R(A)
W(A)

R(A)
W(A)

T1
T2

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.19

 T1 T2
R(A)

R(A)
W(A)
W(A)

T1--->T2
OR
T1 T2--->T1
T2

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.20

 T1 T2 T3
R(A)

R(A)
W(A)

W(A)

R(A)
W(A)

W(A)

T1--->T2--->T3
T1---->T3---->T2 Factorial of no. of Conflict
T2---->T1----->T3 tractions View
T2----->T3---->T1 3!=6
T3----->T1----->T2
T3----->T2------>T1
José Alferes - Adaptado de Database System Concepts - 5
th
Edition 15-17.21
 T1 T2 T3
R(X)

R(Y)
R(X)
R(Y)
R(Z)

W(Y)

W(Z)

R(Z)
W(X)
W(Z)

R(X)--->W(X) Check conflicting pairs in other

W(X)---->W(X) transaction
W(X)-----> R(X)
José Alferes - Adaptado de Database System Concepts - 5
th
Edition 15-17.22
 Indegree=0
T1 T2
T2--->T3--->T1

T3

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.23

 Conflict Equivalent Schedules
S S’
T1 T2 T1 T2
R(A) R(A)
W(A) W(A)
R(A) R(B)
W(A) R(A)
R(B) W(A)

R(A) --> R(A) Non-Conflicting Pair If S=S’

R(A) --> W(A) then S is Conflict Serializable
W(A) --> W(A) Conflicting Pair
W(A) --> R(A)
R(B) --> R(A)
W(B) --> R(A)
Non-Conflicting Pair
R(B) --> W(A)
W(A) ---> W(B)

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.24

 S S
T1 T2 T1 T2
R(A) R(A)
W(A) W(A)
R(A) R(A)
W(A) R(B)
R(B) W(A)

Swap these two operations in Table1 Swapped-Table2

T1 T2 T1 T2
R(A) R(A)
W(A) W(A)
R(A) R(B)
R(B) R(A)
W(A) W(A)

Swap these two operations in Table 2 Final Result equivalent to S’

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.25
 View Serializability

Inconsistent Conflict Serializable

Consistent
Yes No
View Blind Write

CS Yes No
Check Not View Serializability
Dependency
graph

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.26

 T1 T2 T1 T2
R(A) R(A)
Initial W(A) W(A)
Read R(A) R(B)
W(A) W(B)
R(B) R(A)
W(B) W(B)
R(B) R(B)
W(B) W(B)

Final
Write

For S--->VS, we need to check whether it is equivalent with any posssible

serial execution schedule

Properties of View Equivalence

• Initial Read
• Final Write
• Update Read
José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.27
 S
T1 T2 T3 T1 T2 T3
R(A) R(A)
W(A) W(A)
W(A) W(A)
W(A) W(A)

1-2-3
T1 1-3-2
T2 2-1-3
2-3-1
3-1-2
T3 3-2-1

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.28

 Serializability
 Goal : Deal with concurrent schedules that are equivalent to some
serial execution:
 Basic Assumption – Each transaction preserves database
consistency.
 Thus serial execution of a set of transactions preserves database
consistency.
 A (possibly concurrent) schedule is serializable if it is equivalent to a
serial schedule. Different forms of schedule equivalence give rise to
the notions of:
1. conflict serializability
2. view serializability
 Simplified view of transactions
 We ignore operations other than read and write instructions
 We assume that transactions may perform arbitrary computations
on data in local buffers in between reads and writes.
 Our simplified schedules consist of only read and write
instructions.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.29

 Conflicting Instructions
 Instructions li and lj of transactions Ti and Tj respectively, conflict if
and only if there exists some item Q accessed by both li and lj, and at
least one of these instructions wrote Q.
1. li = read(Q), lj = read(Q). li and lj don’t conflict.
2. li = read(Q), lj = write(Q). They conflict.
3. li = write(Q), lj = read(Q). They conflict
4. li = write(Q), lj = write(Q). They conflict
 Intuitively, a conflict between li and lj forces an order between them.
 If li and lj are consecutive in a schedule and they do not conflict,
their results would remain the same even if they had been
interchanged in the schedule.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.30

 Conflict Serializability
 If a schedule S can be transformed into a schedule S´ by a series of swaps of
non-conflicting instructions, we say that S and S´ are conflict equivalent.
 We say that a schedule S is conflict serializable if it is conflict equivalent to a
serial schedule
 Schedule 3 can be transformed into Schedule 6, a serial schedule where T2
follows T1, by series of swaps of non-conflicting instructions. Therefore it is
conflict serializable.

Schedule 3 Schedule 6
José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.31
 Conflict Serializability (Cont.)

 Example of a schedule that is not conflict serializable:

 We are unable to swap instructions in the above schedule to obtain

either the serial schedule < T3, T4 >, or the serial schedule < T4, T3 >.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.32

 Testing for Serializability
 Consider some schedule of a set of transactions T1, T2, ..., Tn
 Precedence graph — a direct graph where
 the vertices are the transactions (names).
 there is an arc from Ti to Tj if the two transaction conflict,
and Ti accessed the data item on which the conflict arose
earlier.
 We may label the arc by the item that was accessed.
 Example 1
x

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.33

 Example Schedule (Schedule A) + Precedence Graph

T1 T2 T3 T4 T5
read(X)
read(Y)
read(Z)
read(V)
read(W) T1 T2
read(W)
read(Y)
write(Y)
write(Z)
read(U)
read(Y)
T3 T4
write(Y)
read(Z)
write(Z)
read(U)
write(U) T5

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.34

 Test for Conflict Serializability
 A schedule is conflict serializable if and only
if its precedence graph is acyclic.
 Cycle-detection algorithms exist which take
order n2 time, where n is the number of
vertices in the graph.
 (Better algorithms take order n + e
where e is the number of edges.)
 If precedence graph is acyclic, the
serializability order can be obtained by a
topological sorting of the graph.
 This is a linear order consistent with the
partial order of the graph.
 For example, a serializability order for
Schedule A would be
T5  T 1  T3  T 2  T4

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.35

 View Serializability
 Sometimes it is possible to serialize schedules that are not conflict
serializable
 View serializability provides a weaker and still consistency preserving
notion of serialization
 Let S and S´ be two schedules with the same set of transactions. S
and S´ are view equivalent if the following three conditions are met,
for each data item Q,
1. If in schedule S, transaction Ti reads the initial value of Q, then in
schedule S’ also transaction Ti must read the initial value of Q.
2. If in schedule S transaction Ti executes read(Q), and that value
was produced by transaction Tj (if any), then in schedule S’ also
transaction Ti must read the value of Q that was produced by the
same write(Q) operation of transaction Tj .
3. The transaction (if any) that performs the final write(Q) operation
in schedule S must also perform the final write(Q) operation in
schedule S’.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.36

 View Serializability (Cont.)
 A schedule S is view serializable if it is view equivalent to a serial
schedule.
 Every conflict serializable schedule is also view serializable.
 Below is a schedule which is view-serializable but not conflict
serializable.

 It is equivalent to either <T3,T4,T6> or <T4,T3,T6>

 Every view serializable schedule that is not conflict serializable has
blind writes.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.37

 Test for View Serializability
 The precedence graph test for conflict serializability cannot be used
directly to test for view serializability.
 Extension to test for view serializability has cost exponential in the
size of the precedence graph.
 The problem of checking if a schedule is view serializable falls in the
class of NP-complete problems.
 Thus existence of an efficient algorithm is extremely unlikely.
 However practical algorithms that just check some sufficient
conditions for view serializability can still be used.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.38

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.39
 Recoverability in DBMS

 As discussed, a transaction may not execute completely due to

hardware failure, system crash or software issues.
 In that case, we have to roll back the failed transaction. But some
other transaction may also have used values produced by the failed
transaction. So we have to roll back those transactions as well.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.40

 Recoverable Schedules
T1 T2 Schedules in which transactions commit
only after all transactions whose changes
R(A) they read commit are called recoverable
A=A-5
W(A) R(A) schedules. In other words, if some
A=A-2 transaction Tj is reading value updated or
W(A) written by some other transaction Ti, then
commit the commit of Tj must occur after the
commit commit of Ti.
Recoverable schedule — if a transaction Tj
reads a data item previously written by a
transaction Ti , then the commit operation of
Ti must appear before the commit operation
of Tj.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.41

 Recoverable Schedules

 Example 1:
 S1: R1(x), W1(x), R2(x), R1(y), R2(y), W2(x), W1(y), C1, C2;

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.42

 Irrecoverable Schedules
T1 T2 Suppose A=10
A=10-5=5
R(A) A=5
A=A-5
W(A) R(A)
A=A-2 A=5
Rollback W(A) A=5-2=3
commit A=3

B=20(Suppose)
R(B)
*
Fail At this
point,Transaction
failed due to any
reason

A=10 again

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.43

 Cascading Rollbacks
 Cascading rollback – a single transaction failure leads to a
series of transaction rollbacks. Consider the following schedule
where none of the transactions has yet committed (so the
schedule is recoverable)

If T10 fails, T11 and T12 must also be rolled back.

 Can lead to the undoing of a significant amount of work

Disadvantage: CPU cycle gets wasted

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.44

 Cascading Rollbacks

 The table below shows a T1 T2

schedule with two transactions,
T1 reads and writes A and that R(A)
value is read and written by T2. A=A-5
But later on, T1 fails. So we W(A) R(A)
have to rollback T1. Since T2 A=A+2
has read the value written by W(A)
T1, it should also be rollbacked. Fail
As it has not committed, we can Commit
rollback T2 as well. So it is Commit
recoverable with cascading
rollback. Therefore, if Tj is
reading value updated by Ti
and commit of Tj is delayed till
commit of Ti, the schedule is
called recoverable with
cascading rollback.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.45

 Cascadeless Schedules
 Cascadeless schedules — in these, cascading rollbacks cannot
occur; for each pair of transactions Ti and Tj such that Tj reads a data
item previously written by Ti, the commit operation of Ti appears
before the read operation of Tj.
 Every cascadeless schedule is also recoverable
 It is desirable to restrict the schedules to those that are cascadeless

Cascadeless Recoverable Rollback:

The table below shows a schedule with two transactions, T1

reads and writes A and commits and that value is read by T2.
But if T1 fails before commit, no other transaction has read its
value, so there is no need to rollback other transaction. So this
is a Cascadeless recoverable schedule. So, if Tj reads value
updated by Ti only after Ti is committed, the schedule will be
cascadeless recoverable.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.46

 Cascadless Rollback

T1 T2

R(A)
A=A-5
W(A)
Commit
R(A)
A=A+2
W(A)
Commit

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.47

  Which of the following scenarios may lead to an irrecoverable error in
a database system?

 A transaction writes a data item after it is read by an uncommitted

transaction.
 A transaction reads a data item after it is read by an uncommitted
transaction.
 A transaction reads a data item after it is written by a committed
transaction.
 A transaction reads a data item after it is written by an uncommitted
transaction.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.48

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.49
José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.50
José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.51
José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.52
 Concurrency Control
 A database must provide a mechanism that will ensure that all possible
schedules are
 either conflict or view serializable, and
 are recoverable and preferably cascadeless
 A policy in which only one transaction can execute at a time generates
serial schedules, but provides a poor degree of concurrency
 Are serial schedules recoverable/cascadeless?
 Testing a schedule for serializability after it has executed is a little too
late!
 Goal – to develop concurrency control protocols that will assure
serializability.
 Lock-based protocols
 Timestamp-based protocols

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.53

 Concurrency Control vs. Serializability Tests

 Concurrency-control protocols allow concurrent schedules, but ensure

that the schedules are conflict/view serializable, and are recoverable
and cascadeless .
 Concurrency control protocols generally do not examine the
precedence graph as it is being created
 Instead a protocol imposes a discipline that avoids nonseralizable
schedules.
 Different concurrency control protocols provide different tradeoffs
between the amount of concurrency they allow and the amount of
overhead that they incur.
 Tests for serializability help us understand why a concurrency control
protocol is correct.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.54

 Lock-Based Protocols
 A lock is a mechanism to control concurrent access to a data item
 Data items can be locked in two modes :
1. exclusive (X) mode. Data item can be both read as well as
written. X-lock is requested using lock-X instruction.
2. shared (S) mode. Data item can only be read. S-lock is
requested using lock-S instruction.
 Lock requests are made to concurrency-control manager. Transaction can
proceed only after request is granted.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.55

 Lock-Based Protocols (Cont.)
 Lock-compatibility matrix

 A transaction may be granted a lock on an item if the requested lock is

compatible with locks already held on the item by other transactions
 Any number of transactions can hold shared locks on an item,
 but if any transaction holds an exclusive on the item no other
transaction may hold any lock on the item.
 If a lock cannot be granted, the requesting transaction is made to wait till
all incompatible locks held by other transactions have been released.
The lock is then granted.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.56

 Lock-Based Protocols (Cont.)
 Example of a transaction performing locking:
T2: lock-S(A);
read (A);
unlock(A);
lock-S(B);
read (B);
unlock(B);
display(A+B)
 Locking as above is not sufficient to guarantee serializability — if A and B
get updated in-between the read of A and B, the displayed sum would be
wrong.
 A locking protocol is a set of rules followed by all transactions while
requesting and releasing locks. Locking protocols restrict the set of
possible schedules.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.57

 The Two-Phase Locking Protocol
 This is a protocol which ensures conflict-serializable schedules.
 Phase 1: Growing Phase
 transaction may obtain locks
 transaction may not release locks
 Phase 2: Shrinking Phase
 transaction may release locks
 transaction may not obtain locks
 The protocol assures serializability. It can be proved that the
transactions can be serialized in the order of their lock points (i.e.
the point where a transaction acquired its final lock).

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.58

 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.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.59

 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.
 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.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.60

 Deadlock Detection
 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.

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

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.61
 Properties of the Two-Phase Locking Protocol

 Cascading roll-back is possible under two-phase locking. To avoid this,

follow a modified protocol called strict two-phase locking. Here a
transaction must hold all its exclusive locks till it commits/aborts.
 Rigorous two-phase locking is even stricter: here all locks are held
till commit/abort. In this protocol transactions can be serialized in the
order in which they commit.
 There can be conflict serializable schedules that cannot be obtained if
two-phase locking is used.
 However, in the absence of extra information (e.g., ordering of access
to data), two-phase locking is needed for conflict serializability in the
following sense:
 Given a transaction Ti that does not follow two-phase locking, we
can find a transaction Tj that uses two-phase locking, and a
schedule for Ti and Tj that is not conflict serializable.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.62

 Multiple Granularity
 Up to now we have considered locking (and execution) at the level of a single
item/row
 However there are circumstances at which it is preferable to perform lock at
different level (sets of tuples, relation, or even sets of relations)
 As extreme example consider a transaction that needs to access to whole
database: performing locks tuple by tuple would be time-consuming
 Allow data items to be of various sizes and define a hierarchy (tree) of data
granularities, where the small granularities are nested within larger ones
 When a transaction locks a node in the tree explicitly, it implicitly locks all the
node's descendents in the same mode.
 Granularity of locking (level in tree where locking is done):
 fine granularity (lower in tree): high concurrency, high locking overhead
 coarse granularity (higher in tree): low locking overhead, low concurrency

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.63

 Example of Granularity Hierarchy

The levels, starting from the coarsest (top) level are

 database
 area
 file
 record

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.64

 Timestamp-Based Protocols
 Instead of determining the order of each operation in a transaction at
execution time, determines the order by the time of beginning of each
transaction.
 Each transaction is issued a timestamp when it enters the system. If an
old transaction Ti has time-stamp TS(Ti), a new transaction Tj is assigned
time-stamp TS(Tj) such that TS(Ti) <TS(Tj).
 The protocol manages concurrent execution such that the time-stamps
determine the serializability order.
 In order to assure such behavior, the protocol maintains for each data Q two
timestamp values:
 W-timestamp(Q) is the largest time-stamp of any transaction that
executed write(Q) successfully.
 R-timestamp(Q) is the largest time-stamp of any transaction that
executed read(Q) successfully.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.65

 Timestamp-Based Protocols (Cont.)
 The timestamp ordering protocol ensures that any conflicting read and write
operations are executed in timestamp order.
 Suppose a transaction Ti issues a read(Q)
1. If TS(Ti)  W-timestamp(Q), then Ti needs to read a value of Q that was
already overwritten.
 Hence, the read operation is rejected, and Ti is rolled back.
2. If TS(Ti) W-timestamp(Q), then the read operation is executed, and R-
timestamp(Q) is set to max(R-timestamp(Q), TS(Ti)).
 Suppose that transaction Ti issues write(Q).
1. If TS(Ti) < R-timestamp(Q), then the value of Q that Ti is producing was
needed previously, and the system assumed that that value would never
be produced.
 Hence, the write operation is rejected, and Ti is rolled back.
2. If TS(Ti) < W-timestamp(Q), then Ti is attempting to write an obsolete
value of Q.
 Hence, this write operation is rejected, and Ti is rolled back.
3. Otherwise, the write operation is executed, and W-timestamp(Q) is set to
TS(Ti).

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.66

 Correctness of Timestamp-Ordering Protocol

 The timestamp-ordering protocol guarantees serializability since all the

arcs in the precedence graph are of the form:

transaction transaction
with smaller with larger
timestamp timestamp

Thus, there will be no cycles in the precedence graph

 Timestamp protocol ensures freedom from deadlock as no transaction
ever waits.
 But the schedule may not be cascade-free, and may not even be
recoverable.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.67

 Multiversion Schemes
 Up to now we only considered a single copy (the most recent) of each
database item.
 Multiversion schemes keep old versions of data item to increase concurrency.
 Multiversion Timestamp Ordering
 Multiversion Two-Phase Locking
 Basic Idea of multiversion schemes
 Each successful write results in the creation of a new version of the data
item written.
 Use timestamps to label versions.
 When a read(Q) operation is issued, select an appropriate version of Q
based on the timestamp of the transaction, and return the value of the
selected version.
 reads never have to wait as an appropriate version is returned
immediately.
 A drawback is that creation of multiple versions increases storage overhead
 Garbage collection mechanism may be used…

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.68

 Multiversion Timestamp Ordering
 Each data item Q has a sequence of versions <Q1, Q2,...., Qm>. Each
version Qk contains three data fields:
 Content -- the value of version Qk.
 W-timestamp(Qk) -- timestamp of the transaction that created
(wrote) version Qk
 R-timestamp(Qk) -- largest timestamp of a transaction that
successfully read version Qk
 when a transaction Ti creates a new version Qk of Q, Qk's W-
timestamp and R-timestamp are initialized to TS(Ti).
 R-timestamp of Qk is updated whenever a transaction Tj reads Qk, and
TS(Tj) > R-timestamp(Qk).

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.69

 Multiversion Timestamp Ordering (Cont)
 Suppose that transaction Ti issues a read(Q) or write(Q) operation. Let
Qk denote the version of Q whose write timestamp is the largest write
timestamp less than or equal to TS(Ti).
1. If transaction Ti issues a read(Q), then the value returned is the
content of version Qk.
2. If transaction Ti issues a write(Q)
1. if TS(Ti) < R-timestamp(Qk), then transaction Ti is rolled back.
2. if TS(Ti) = W-timestamp(Qk), the contents of Qk are overwritten
3. else a new version of Q is created.
 Observe that
 Reads always succeed
 A write by Ti is rejected if some other transaction Tj that (in the
serialization order defined by the timestamp values) should read
Ti's write, has already read a version created by a transaction older
than Ti.
 Protocol guarantees serializability
José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.70
 Multiversion Two-Phase Locking
 Differentiates between read-only transactions and update transactions
 Update transactions acquire read and write locks, and hold all locks up
to the end of the transaction. That is, update transactions follow rigorous
two-phase locking.
 Each successful write results in the creation of a new version of the
data item written.
 each version of a data item has a single timestamp whose value is
obtained from a counter ts-counter that is incremented during
commit processing.
 Read-only transactions are assigned a timestamp by reading the current
value of ts-counter before they start execution; they follow the
multiversion timestamp-ordering protocol for performing reads.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.71

 Multiversion Two-Phase Locking (Cont.)
 When an update transaction wants to read a data item:
 it obtains a shared lock on it, and reads the latest version.
 When it wants to write an item
 it obtains X lock on; it then creates a new version of the item and
sets this version's timestamp to .
 When update transaction Ti completes, commit processing occurs:
 Ti sets timestamp on the versions it has created to ts-counter + 1
 Ti increments ts-counter by 1
 Read-only transactions that start after Ti increments ts-counter will see
the values updated by Ti.
 Read-only transactions that start before Ti increments the
ts-counter will see the value before the updates by Ti.
 Only serializable schedules are produced.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.72

 Weak Levels of Consistency
 Some applications are willing to live with weak levels of consistency,
allowing schedules that are not serializable
 E.g. a read-only transaction that wants to get an approximate total
balance of all accounts
 E.g. database statistics computed for query optimization can be
approximate
 Such transactions need not be serializable with respect to other
transactions
 Tradeoff accuracy for performance

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.73

 Levels of Consistency in SQL-92
 Serializable — default in SQL standard
 Repeatable read — only committed records to be read, repeated
reads of same record must return same value (no updates of read
items in between). However, a transaction may not be serializable – it
may find some records inserted by a transaction but not find others.
 Read committed — only committed records can be read, but
successive reads of record may return different (but committed)
values.
 Read uncommitted — even uncommitted records may be read.

 In many database systems, such as Oracle, read committed

is the default consistency level
 has to be explicitly changed to serializable when required
 set isolation level serializable
 Lower degrees of consistency useful for gathering
approximate
information
José Alferes - Adaptado about
de Database System the
Concepts database
- 5 Edition
th
15-17.74
 Recovery Schemes
 Recovery schemes are techniques to ensure database consistency
and transaction atomicity and durability despite failures such as
transaction failures, system crashes, disk failures.
 We just briefly focus this issue, which strongly relies on lower-
level control (usage of RAID, buffer management)
 More on this can be found in chapter 17 of the book
 Recovery algorithms have two parts
1. Actions taken during normal transaction processing to ensure
enough information exists to recover from failures
2. Actions taken after a failure to recover the database contents to a
state that ensures atomicity, consistency and durability

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.75

 Recovery and Atomicity
 Modifying the database without ensuring that the transaction commits
may leave the database in an inconsistent state.
 Consider again the transaction Ti that transfers €50 from account A
to account B.
 Several output operations are required for Ti (to output A and B). A
failure may occur after one of these modifications have been made
but before all of them are made.
 To ensure atomicity despite failures, first output information describing
the modifications to stable storage (i.e. storage guaranteed/assumed not
to fail, e.g. with RAID) without modifying the database itself.
 Two approaches are possible:
 log-based recovery, and
 shadow-paging

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.76

 Log-Based Recovery
 A log is kept on stable storage.
 The log is a sequence of log records, and maintains a record of
update activities on the database.
 When transaction Ti starts, it registers itself by writing a
<Ti start>log record
 Before Ti executes write(X), a log record <Ti, X, V1, V2> is written,
where V1 is the value of X before the write, and V2 is the value to be
written to X.
 Log record notes that Ti has performed a write on data item Xj Xj
had value V1 before the write, and will have value V2 after the write.
 When Ti finishes it last statement, the log record <Ti commit> is written.
 For writing the actual records
 Deferred database modification
 Immediate database modification

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.77

 Deferred Database Modification
 The deferred database modification scheme records all
modifications to the log, and defers writes to after partial commit.
 Transaction starts by writing <Ti start> record to log.
 A write(X) operation results in a log record <Ti, X, V> being written,
where V is the new value for X (old value is not needed).
 The write is not performed on X at this time, but is deferred.
 When Ti partially commits, <Ti commit> is written to the log
 After that, the log records are read and used to actually execute the
previously deferred writes.
 During recovery after a crash, a transaction needs to be redone if and
only if both <Ti start> and<Ti commit> are there in the log.
 Redoing a transaction Ti ( redoTi) sets the value of all data items
updated by the transaction to the new values.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.78

 Immediate Database Modification
 The immediate database modification scheme allows database
updates of an uncommitted transaction to be made as the writes are
issued
 since undoing may be needed, update logs must have both old
value and new value
 Update log record must be written before database item is written
 We assume that the log record is output directly to stable storage
 Can be extended to postpone log record output, so long as prior to
execution of an output(B) operation for a data block B, all log
records corresponding to items B must be flushed to stable
storage
 Output of updated blocks can take place at any time before or after
transaction commit
 Order in which blocks are output can be different from the order in
which they are written.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.79

 Immediate Database Modification (Cont.)
 Recovery procedure has two operations instead of one:
 undo(Ti) restores the value of all data items updated by Ti to their
old values, going backwards from the last log record for Ti
redo(Ti) sets the value of all data items updated by Ti to the new

values, going forward from the first log record for Ti
 Both operations must be idempotent
 That is, even if the operation is executed multiple times the effect is
the same as if it is executed once
 Needed since operations may get re-executed during recovery
 When recovering after failure:
 Transaction Ti needs to be undone if the log contains the record
<Ti start>, but does not contain the record <Ti commit>.
Transaction Ti needs to be redone if the log contains both the record

<Ti start> and the record <Ti commit>.
 Undo operations are performed first, then redo operations.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.80

 Checkpoints
 Problems in recovery procedure as discussed earlier :
1. searching the entire log is time-consuming
2. one might unnecessarily redo transactions which have already
output their updates to the database.
 Streamline recovery procedure by periodically performing
checkpointing
1. Output all log records currently residing in main memory onto
stable storage.
2. Output all modified buffer blocks to the disk.
3. Write a log record < checkpoint> onto stable storage.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.81

 Shadow Paging
 Shadow paging is an alternative to log-based recovery; this scheme is
useful if transactions execute serially
 Idea: maintain two page tables during the lifetime of a transaction –the
current page table, and the shadow page table
 Store the shadow page table in nonvolatile storage, such that state of the
database prior to transaction execution may be recovered.
 Shadow page table is never modified during execution
 To start with, both the page tables are identical. Only current page table is
used for data item accesses during execution of the transaction.
 Whenever any page is about to be written for the first time
 A copy of this page is made onto an unused page.
 The current page table is then made to point to the copy
 The update is performed on the copy

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.82

 Transaction Definition in SQL
 Data manipulation language must include a construct for
specifying the set of actions that comprise a transaction.
 In SQL, a transaction begins implicitly, after previous transaction.
 A transaction in SQL ends by:
 Commit work commits current transaction and begins a new
one.
 Rollback work causes current transaction to abort.
 In almost all database systems, by default, every SQL statement
also commits implicitly if it executes successfully
 Implicit commit can be turned off by a database directive
 E.g. in JDBC, connection.setAutoCommit(false);
 Four levels of (weak) consistency, cf. before.

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.83

 Transaction management in Oracle
 Transaction beginning and ending as in SQL
 Explicit commit work and rollback work
 Implicit commit on session end, and implicit rollback on failure
 Log-based deferred recovery using rollback segment
 Checkpoints (inside transactions) can be handled explicitly
 savepoint <name>
 rollback to <name>

 Concurrency control is made by (a variant of) multiversion rigorous

two-phase locking
 Deadlock are detected using a wait-graph
 Upon deadlock detection, the last transaction that detects the
deadlock is rolled back

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.84

 Levels of Consistency in Oracle
 Oracle implements 2 of the 4 of levels of SQL
 Read committed, by default in Oracle and with
 set transaction isolation level read committed
 Serializable, with
 set transaction isolation level serializable
 Appropriate for large databases with only few updates, and
usually with not many conflicts. Otherwise it is too costly.
 Further, it supports a level similar to repeatable read:
 Read only mode, only allow reads on committed data, and further
doesn’t allow INSERT, UPDATE or DELETE on that data. (without
unrepeatable reads!)
 set transaction isolation level read only

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.85

 Granularity in Oracle
 By default Oracle performs row level locking.
 Command
select … for update
locks the selected rows so that other users cannot lock or update the
rows until you end your transaction. Restriction:
 Only at top-level select (not in sub-queries)
 Not possible with DISTINCT operator, CURSOR expression, set
operators, group by clause, or aggregate functions.
 Explicit locking of tables is possible in several modes, with
 lock table <name> in
 row share mode
 row exclusive mode
 share mode
 share row exclusive mode
 exclusive mode

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.86

 Lock modes in Oracle
 Row share mode
 The least restrictive mode (with highest degree of concurrency)
 Allows other transactions to query, insert, update, delete, or lock
rows concurrently in the same table, except for exclusive mode
 Row exclusive mode
 As before, but doesn’t allow setting other modes except for row
share.
 Acquired automatically after a insert, update or delete command
on a table
 Exclusive mode
 Only allows queries to records of the locked table
 No modifications are allowed
 No other transaction can lock the table in any other mode

 See manual for details of other (intermediate) modes

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.87

 Consistency tests in Oracle
 By default, in Oracle all consistency tests are made immediately after each
DML command (insert, delete or update).
 However, it is possible to defer consistency checking of constraints (primary
keys, candidate keys, foreign keys, and check conditions) to the end of
transactions.
 Only this makes it possible e.g. to insert tuples in relation with circular
dependencies in foreign keys
 For this:
 each constraints that may possibly be deferred must be declared as
deferrable:
 At the definition of the constraint add deferrable immediately
afterwards
 at the transaction in which one wants to defer the verification of the
constraints, add command:
 set constraints all deferred
 In this command, instead of all it is possible to specify which
constraints are to be deferred, by giving their names separated by
commas

José Alferes - Adaptado de Database System Concepts - 5th Edition 15-17.88