By Purvi Gautam

Transaction Management
Transaction management
 The nice aspect of this historical perspective is that it does
indeed encompass some of the fundamental properties of
a transaction (atomicity and durability) as the term is used
in database systems.
 A transaction takes a database, performs an action on it,
and generates a new version of the database, causing a
state transition.
 In general, a transaction is considered to be made up of a
sequence of read and write operations on the database,
together with computation steps.
 Another definition of a transaction is that it is a single
execution of a program .
Properties of transaction

 The consistency and reliability aspects of

transactions are due to four properties:
 (1) atomicity
 (2) consistency
 (3) isolation
 (4) durability
 these are commonly referred to as the ACID
properties of transactions.
Atomicity
 Atomicity refers to the fact that a transaction is treated as a unit of
operation.
 Therefore, either all the transaction’s actions are completed, or none of
them are.
 Atomicity requires that if the execution of a transaction is interrupted by
any sort of failure, the DBMS will be responsible for determining what to
do with the transaction upon recovery from the failure.
 Two possible courses of action: it can either be terminated by completing
the remaining actions, or it can be terminated by undoing all the actions
that have already been executed.
 A transaction itself may fail due to input data errors, deadlocks, or other
factors is known as transaction recovery.
 The second type of failure is caused by system crashes, such as media
failures, processor failures, communication link breakages, power
outages, and so on i.e known as crash recovery.
Consistency

 The consistency of a transaction is simply its

correctness.
 A transaction is a correct program that maps
one consistent database state to another.
 The point in defining multiple levels of
consistency is to provide application
programmers the flexibility to define
transactions that operate at different levels.
Isolation
 Isolation is the property of transactions that requires
each transaction to see a consistent database at all
times.
 An executing transaction cannot reveal its results to
other concurrent transactions before its commitment.
 There are a number of reasons for insisting on
isolation.
 One has to do with maintaining the interconsistency of
transactions.
 If two concurrent transactions access a data item that
is being updated by one of them, it is not possible to
guarantee that the second will read the correct value.
Durability

 Durability refers to that property of transactions

which ensures that once a transaction commits,
its results are permanent and cannot be erased
from the database.
 the DBMS ensures that the results of a transaction
will survive subsequent system failures.
 The durability property brings forth the issue of
database recovery, that is, how to recover the
database to a consistent state where all the
committed actions are reflected.
Types of Transactions

 Flat Transactions
 Nested Transactions
 Workflows
Human-oriented workflows,
System-oriented workflows
Transactional workflows
Flat transactions& Nested
Transactions
 Flat transactions have a single start point (Begin
transaction) and a single termination point (End
transaction).
 Most of the transaction management work in
databases has concentrated on flat transactions.
 An alternative transaction model is to permit a
transaction to include other transactions with their
own begin and commit points. Such transactions are
called nested transactions.
 These transactions that are embedded in another one
are usually called subtransactions.
Contd..
 Nested transactions have received considerable interest as a
more generalized transaction concept.
 The level of nesting is generally open, allowing
subtransactions themselves to have nested transactions.
 The advantages of nested transactions are the following:
 First, they provide a higher-level of concurrency among
transactions.
 A second argument in favor of nested transactions is related
to recovery.
 Finally, it is possible to create new transactions from
existing ones simply by
 inserting the old one inside the new one as a subtransaction.
Workflows

 The term “workflow,” unfortunately, does not

have a clear and uniformly accepted
meaning. A working definition is that a
workflow is “a collection of tasks organized to
accomplish some business process.”
 This definition, however, leaves a lot
undefined.
 Three types of workflows are identified:
Contd..
 Human-oriented workflows, which involve humans in
performing the tasks.
 The system support is provided to facilitate collaboration
and coordination among humans, but it is the humans
themselves who are ultimately responsible for the
consistency of the actions.
 System-oriented workflows are those that consist of
computation-intensive and specialized tasks that can be
executed by a computer.
 The system support in this case is substantial and involves
concurrency control and recovery, automatic task
execution, notification, etc.
Contd..
 Transactional workflows range in between human-
oriented and system oriented workflows and borrow
characteristics from both.
 They involve “coordinated execution of multiple tasks
that
 (a) may involve humans
 (b) require access to HAD [heterogeneous,
autonomous, and/or distributed] systems, and
 (c) support selective use of transactional properties
[i.e., ACID properties] for
 individual tasks or entire workflows.”
Distributed Concurrency
Control
Serializability Theory
 if the concurrent execution of transactions leaves the
database in a state that can be achieved by their serial
execution in some order, problems such as lost
updates will be resolved.
 This is exactly the point of the serializability argument.
 we can indicate that the primary function of a
concurrency controller is to generate a serializable
history for the execution of pending transactions.
 Serializability theory extends in a straightforward
manner to the non-replicated (or partitioned)
distributed databases.
Concurrency control mechanism &
algorithm
 The concurrency control mechanisms into
two broad classes: pessimistic concurrency
control methods and optimistic concurrency
control methods.
 Two methods of concurrency control:
Algorithm
Contd..

 In the locking-based approach, the

synchronization of transactions is achieved
by employing physical or logical locks on
some portion or granule of the database.
 The size of these portions (usually called
locking granularity) is an important issue.
 However, for the time being, we will ignore it
and refer to the chosen granule as a lock unit.
Contd.
 The timestamp ordering (TO) class involves
organizing the execution order of transactions so that
they maintain transaction consistency.
 This ordering is maintained by assigning timestamps
to both the transactions and the data items that are
stored in the database.
 We should indicate that in some locking-based
algorithms, timestamps are also used. This is done
primarily to improve efficiency and the level of
concurrency.
 We call these hybrid algorithms.
Timestamp-Based Concurrency
Control Algorithms
 Unlike locking-based algorithms, timestamp-based
concurrency control algorithms do not attempt to maintain
serializability by mutual exclusion.
 A timestamp is a simple identifier that serves to identify each
transaction uniquely and is used for ordering.
 Uniqueness is only one of the properties of timestamp
generation. The second property is monotonicity.
 Two timestamps generated by the same transaction manager
should be monotonically increasing. Thus timestamps are
values derived from a totally ordered domain.
 There are a number of ways that timestamps can be
assigned:
Contd.

 One method is to use a global (system-wide)

monotonically increasing counter.
 To maintain uniqueness, each site appends its
own identifier to the counter value.
Optimistic Concurrency Control
Algorithms
 The conflicts between transactions are quite
frequent and do not permit a transaction to
access a data item if there is a conflicting
transaction that accesses that data item.
 The execution of any operation of a transaction
follows the sequence of phases: validation (V),
read (R), computation (C), write (W). Generally,
this sequence is valid for an update transaction
as well as for each of its operations.
Contd..
 Delay the validation phase until just before the write
phase. Thus an operation submitted to an optimistic
scheduler is never delayed.
 The read, compute, and write operations of each
transaction are processed freely without updating the
actual database.
 The validation phase consists of checking if these updates
would maintain the consistency of the database.
 If the answer is affirmative, the changes are made global
(i.e., written into the actual database). Otherwise, the
transaction is aborted and has to restart.
Deadlock Management
 Any locking-based concurrency control algorithm may
result in deadlocks, since there is mutual exclusion of
access to shared resources (data) and transactions may
wait on locks.
 A deadlock can occur because transactions wait for one
another.
 Informally, a deadlock situation is a set of requests that
can never be granted by the concurrency control
mechanism.
 A deadlock is a permanent phenomenon. If one exists in a
system, it will not go away unless outside intervention
takes place.
Contd..

 A useful tool in analyzing deadlocks is a wait-for

graph (WFG).
 WFG is a directed graph that represents the wait-
for relationship among transactions.
 Using the WFG, it is easier to indicate the
condition for the occurrence of a deadlock. A
deadlock occurs when the WFG contains a cycle.
 There are three known methods for handling
deadlocks: prevention, avoidance, and detection
and resolution.
Contd..
 Deadlock Prevention:Deadlock prevention
methods guarantee that deadlocks cannot occur in
the first place. Thus the transaction manager checks
a transaction when it is first initiated and does not
permit it to proceed if it may cause a deadlock.
 The transaction manager then permits a transaction
to proceed if all the data items that it will access are
available.
 Otherwise, the transaction is not permitted
to proceed.
Contd..

 Deadlock Avoidance: Deadlock avoidance schemes

either employ concurrency control techniques that
will never result in deadlocks or require that
potential deadlock situations are detected in
advance and steps are taken such that they will not
occur.
 The simplest means of avoiding deadlocks is to
order the resources and insist that each process
request access to these resources in that order.
 Deadlock avoidance methods are more suitable than
prevention schemes for database environments.
Contd.
 Deadlock Detection and Resolution:Deadlock detection
and resolution is the most popular and best-studied
method.
 The amount of effort that has already been invested in the
transaction. This effort will be lost if the transaction is
aborted.
 The cost of aborting the transaction. This cost generally
depends on the number of updates that the transaction
has already performed.
 The amount of effort it will take to finish executing the
transaction. The scheduler wants to avoid aborting a
transaction that is almost finished.