UNIT IV

TRANSACTION AND
CONCURRENCY

1
Transaction Concept
A transaction is a unit of program execution that
accesses and possibly updates various data items.
E.g., transaction to transfer $50 from account A to
account B:
1. read(A)
2. A := A – 50
3. write(A)
4. read(B)
5. B := B + 50
6. write(B)
Two main issues to deal with:
Failures of various kinds, such as hardware failures
and system crashes
Concurrent execution of multiple transactions
 Required Properties of a
Transaction
 Consider a transaction to transfer $50 from account A to account B:
1. read(A)
2. A := A – 50
3. write(A)
4. read(B)
5. B := B + 50
6. write(B)
 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
 Durability requirement — once the user has been notified that the transaction
has completed (i.e., the transfer of the $50 has taken place), the updates to the
database by the transaction must persist even if there are software or hardware
failures.
Required Properties of a Transaction (Cont.)

 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, when starting to execute, must see a consistent
database.
 During transaction execution the database may be temporarily
inconsistent.
 When the transaction completes successfully the database must be
consistent
 Erroneous transaction logic can lead to inconsistency
 Required Properties of a Transaction (Cont.)
 Isolation requirement — if between steps 3 and 6 (of the fund transfer
transaction) , 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)
2. A := A – 50
3. write(A)
read(A), read(B), print(A+B)
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.
ACID Properties
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 transaction in isolation 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.
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.
Transaction State (Cont.)
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
 Will study in Chapter 15, after studying notion of correctness of
concurrent executions.
 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
 Schedule 1
 Let T1 transfer $50 from A to B, and T2 transfer 10% of the balance from A to B.
 An example of a serial schedule in which T1 is followed by T2 :
 Schedule 2
 A serial schedule in which T2 is followed by T1 :
 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.

Note -- In schedules 1, 2 and 3, the sum “A + B” is preserved.

Schedule 4
 The following concurrent schedule does not preserve
the sum of “A + B”
Serializability
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.
Conflicting Instructions
Let li and lj be two Instructions of transactions Ti and Tj
respectively. Instructions li and lj 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 a (logical)
temporal 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.
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
Conflict Serializability (Cont.)
 Schedule 3 can be transformed into Schedule 6 -- a serial
schedule where T2 follows T1, by a series of swaps of non-
conflicting instructions. Therefore, Schedule 3 is conflict
serializable.

Schedule 3 Schedule 6
Recoverable Schedules
 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.
 The following schedule is not recoverable if T9 commits immediately
after the read(A) operation.

 If T8 should abort, T9 would have read (and possibly shown to the user)
an inconsistent database state. Hence, database must ensure that
schedules are recoverable.
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
Cascadeless Schedules
 Cascadeless schedules — 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
 Example of a schedule that is NOT cascadeless
Concurrency Control
 A database must provide a mechanism that will ensure that all possible
schedules are both:
 Conflict serializable.
 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
 Concurrency-control schemes tradeoff between the amount of
concurrency they allow and the amount of overhead that they incur
 Testing a schedule for serializability after it has executed is a little too
late!
 Tests for serializability help us understand why a concurrency
control protocol is correct
Goal – to develop concurrency control protocols that
will assure serializability.
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 (why?)
Such transactions need not be serializable with
respect to other transactions
Tradeoff accuracy for performance
 Levels of Consistency in SQL-92
 Serializable — default
 Repeatable read — only committed records to be read, repeated reads of
same record must return same value. 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.
 Lower degrees of consistency useful for gathering approximate
information about the database
 Warning: some database systems do not ensure serializable
schedules by default
 E.g., Oracle and PostgreSQL by default support a level of
consistency called snapshot isolation (not part of the SQL
standard)
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.
 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);
Other Notions of Serializability
View Serializability
 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’.
 As can be seen, view equivalence is also based purely on reads and
writes alone.
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.

 What serial schedule is above equivalent to?

 Every view serializable schedule that is not conflict serializable
has blind writes.
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.
More Complex Notions of Serializability
 The schedule below produces the same outcome as the serial schedule <
T1, T5 >, yet is not conflict equivalent or view equivalent to it.

 If we start with A = 1000 and B = 2000, the final result is 960 and 2040
 Determining such equivalence requires analysis of operations other than
read and write.
Failure Classification
 Transaction failure :
 Logical errors: transaction cannot complete due to some internal error
condition
 System errors: the database system must terminate an active transaction due
to an error condition (e.g., deadlock)
 System crash: a power failure or other hardware or software failure causes the
system to crash.
 Fail-stop assumption: non-volatile storage contents are assumed to not be
corrupted by system crash
 Database systems have numerous integrity checks to prevent corruption of
disk data
 Disk failure: a head crash or similar disk failure destroys all or part of disk
storage
 Destruction is assumed to be detectable: disk drives use checksums to detect
failures
 Recovery Algorithms
 Consider transaction Ti that transfers $50 from account A to account B
 Two updates: subtract 50 from A and add 50 to B
 Transaction Ti requires updates to A and B to be output to the database.
 A failure may occur after one of these modifications have been made
but before both of them are made.
 Modifying the database without ensuring that the transaction will
commit may leave the database in an inconsistent state
 Not modifying the database may result in lost updates if failure
occurs just after transaction commits
 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
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 the concurrency-control
manager by the programmer. Transaction can proceed
only after request is granted.
 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.
 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.
 The Two-Phase Locking Protocol
 This protocol 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).
 The Two-Phase Locking Protocol
 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.
 Lock Conversions
Two-phase locking with lock conversions:
– First Phase:
can acquire a lock-S on item
can acquire a lock-X on item
can convert a lock-S to a lock-X (upgrade)
– Second Phase:
can release a lock-S
can release a lock-X
can convert a lock-X to a lock-S (downgrade)
 This protocol assures serializability. But still relies on the
programmer to insert the various locking instructions.
 Deadlocks
 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-S(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.
 Deadlocks (Cont.)
 Two-phase locking does not ensure freedom from deadlocks.
 In addition to deadlocks, there is a possibility of starvation.
 Starvation occurs if the 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.
 Deadlocks (Cont.)
 The potential for deadlock exists in most locking protocols.
Deadlocks are a necessary evil.
 When a deadlock occurs there is a possibility of cascading roll-
backs.
 Cascading roll-back is possible under two-phase locking. To
avoid this, follow a modified protocol called strict two-phase
locking -- 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.
 Implementation of Locking
A lock manager can be implemented as a separate process to
which transactions send lock and unlock requests
The lock manager replies to a lock request by sending a lock
grant messages (or a message asking the transaction to roll
back, in case of a deadlock)
The requesting transaction waits until its request is answered
The lock manager maintains a data-structure called a lock
table to record granted locks and pending requests
The lock table is usually implemented as an in-memory hash
table indexed on the name of the data item being locked
 Deadlock Handling
System is deadlocked if there is a set of transactions
such that every transaction in the set is waiting for
another transaction in the set.
Deadlock prevention protocols ensure that the system
will never enter into a deadlock state. Some prevention
strategies :
Require that each transaction locks all its data items
before it begins execution (predeclaration).
Impose partial ordering of all data items and require
that a transaction can lock data items only in the
order specified by the partial order.
 More Deadlock Prevention Strategies
 Following schemes use transaction timestamps for the sake of
deadlock prevention alone.
 wait-die scheme — non-preemptive
older transaction may wait for younger one to release data
item. (older means smaller timestamp) Younger transactions
never Younger transactions never wait for older ones; they
are rolled back instead.
a transaction may die several times before acquiring needed
data item
 wound-wait scheme — preemptive
older transaction wounds (forces rollback) of younger
transaction instead of waiting for it. Younger transactions
may wait for older ones.
may be fewer rollbacks than wait-die scheme.
 Deadlock prevention (Cont.)
 Both in wait-die and in wound-wait schemes, a rolled back
transactions is restarted with its original timestamp. Older
transactions thus have precedence over newer ones, and
starvation is hence avoided.
 Timeout-Based Schemes:
a transaction waits for a lock only for a specified amount of
time. If the lock has not been granted within that time, the
transaction is rolled back and restarted,
Thus, deadlocks are not possible
simple to implement; but starvation is possible. Also difficult
to determine good value of the timeout interval.
Deadlock Detection
 Deadlocks can be described as a wait-for graph, which consists of a
pair G = (V,E),
 V is a set of vertices (all the transactions in the system)
 E is a set of edges; each element is an ordered pair Ti Tj.
 If Ti Tj is in E, then there is a directed edge from Ti to Tj,
implying that Ti is waiting for Tj to release a data item.
 When Ti requests a data item currently being held by Tj, then the
edge Ti  Tj is inserted in the wait-for graph. This edge is removed
only when Tj 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.
Deadlock Detection (Cont.)

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

 Deadlock Recovery
 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.
Rollback -- determine how far to roll back transaction
Total rollback: Abort the transaction and then restart it.
More effective to roll back transaction only as far as
necessary to break deadlock.
Starvation happens if same transaction is always chosen as
victim. Include the number of rollbacks in the cost factor to
avoid starvation
 Recovery and Atomicity
To ensure atomicity despite failures, we first output
information describing the modifications to stable
storage without modifying the database itself.
We study log-based recovery mechanisms in detail
We first present key concepts
And then present the actual recovery algorithm
Less used alternative: shadow-copy and shadow-
paging (brief details in book)
shadow-copy
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 (the old
value), and V2 is the value to be written to X (the new value).
When Ti finishes it last statement, the log record <Ti commit>
is written.
Two approaches using logs
Deferred database modification
Immediate database modification
Immediate Database Modification
 The immediate-modification scheme allows updates of an uncommitted
transaction to be made to the buffer, or the disk itself, before the
transaction commits
 Update log record must be written before database item is written
 We assume that the log record is output directly to stable storage
 (Will see later that how to postpone log record output to some extent)
 Output of updated blocks to stable storage 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.
 The deferred-modification scheme performs updates to buffer/disk only
at the time of transaction commit
 Simplifies some aspects of recovery
 But has overhead of storing local copy
Transaction Commit
A transaction is said to have committed when its
commit log record is output to stable storage
all previous log records of the transaction must have
been output already
Writes performed by a transaction may still be in the
buffer when the transaction commits, and may be
output later
Concurrency Control and Recovery
 With concurrent transactions, all transactions share a single disk buffer
and a single log
 A buffer block can have data items updated by one or more
transactions
 We assume that if a transaction Ti has modified an item, no other
transaction can modify the same item until Ti has committed or aborted
 i.e. the updates of uncommitted transactions should not be visible to
other transactions
 Otherwise how to perform undo if T1 updates A, then T2 updates
A and commits, and finally T1 has to abort?
 Can be ensured by obtaining exclusive locks on updated items and
holding the locks till end of transaction (strict two-phase locking)
 Log records of different transactions may be interspersed in the log.