Twana Latif Mohammed
Transaction is a
fixed of query instructions to being execute atomically on a one constant view
of a database. that is executed with the aid of imposing ACID (Atomicity,
Consistency, Isolation and durability) Characteristics without compromising the
scalability properties of the database. but, the underlying records storage
services offer frequently eventual consistency. then, we’ve got surveyed vital
research literature transaction in DDBMSs 7. (Figure 1 presents the
transaction model 2).
Figure 1: Transaction Model
In the transaction Management when concurrent accesses and
failures occur then they have to do with the problems of ever protection the
database in consistent state even (Figure 2). A transaction is a number of actions
that produce consistent transformations of system states while preserving
system consistency 7.
transaction went to another step in much more complex application like
“distribution, process orientation and loose coupling”. 3 In program database
application the transaction is very important then they resolve the two main
problems for application designer. First is transact
3 Concurrent programming is complex and it is exactly error then
this error not easy to debug and to
remaking. Transaction by
using concurrency control able to solve this problem. The designer to access
the database only need to bracket with transaction that have concurrency
control for example, in two bank account the operation between them.
feature, after developed the transaction should support multiple organizational
field process by activated business network that is performed. On the other
side, provided the transaction to more technical called “ACID” properties. ACID
properties are a transaction can be term that in an atomic way by set of data
operations that are executed.
4 Transaction consist of four proprieties:
This refers to the fact that
a transaction is treated as a unit of operation. Consequently, it dictates that
either all the actions related to a transaction are completed or none of them
is carried out. For example, in the case of a crash, the system should complete
of the transaction, or it will undo all the actions pertaining to
this transaction. The recovery of the transaction is split into two types
corresponding to the two types of failures: the transaction recovery, which is
due to the system terminating one of the transactions because of deadlock
handling; and the crash recovery, which is done after a system crash or a
Referring to its correctness, this property deals with maintaining
consistent data in a database system. Consistency falls under the subject of
For example, dirty data is data that has been
modified by a transaction that has not yet committed. Thus, the job of
concurrency control is to be able to disallow transactions from reading or
updating dirty data.
According to this property,
each transaction should see a consistent database at all times. Consequently,
no other transaction can read or modify data that is being modified by another
transaction. If this property is not maintained, one of two things could happen
to the database.
A. Lost Updates: this occurs when another
transaction (T2) updates the same data being modified by the first transaction
(T1) in such a manner that T2 reads the value prior to the writing of T1 thus
creating the problem of losing this update.
B. Cascading Aborts: this problem occurs when
the first transaction (T1) aborts, then the transactions that had read or
modified data that has been used by T1 will also abort.
This property ensures that
once a transaction commits, its results are permanent and cannot be erased from
the database. This means that whatever happens after the COMMIT of a
transaction, whether it is a system crash or aborts of other transactions, the results
already committed are not modified or undone.
There are many
study in the literature that talking about the techniques for maintenance and
analysis of transaction management in database. In addition, the related works
discuss about the transaction of database integrity models, techniques. In this
literature, they are more scheduling steps and techniques are available but,
largest number of them are based on the models of simple system, which don’t
the inverse of transaction process and cargo balancing.
The authors in this
paper 5 attempted to describe the significant concepts of transaction
management in Distribute Database System (DDBS) and how Oracle implements this
technique. Two-phase commit technique which it’s one of Oracle features is used
to manage the enterprise data resource successfully of any organization via
constructs a session tree for the participating nodes for each transaction. in
any given transaction the session tree describes the relations between the
nodes participating. 1 In this study by advanced and work?ow transactions to
the Web Services and Grid transaction models shows thereby investigating
numerous transaction models ranging from the classical ?at transactions.
As a result,
the area of process and service compositions requires the support for
transaction composition. Also, the Web service transactions and the mobile
transactions might converge as they are both supporting thin, small clients
which represent autonomous ‘parties’ of which the accessibility cannot be
determined beforehand. 3 In this paper how it can be easily integrated in any
versioned data store and focus on the ultra-scalable transaction management
API. An ultra-scalable transactional management layer that can be integrated
with any data store with multi-versioning capabilities. Finally, data can also
be queried across data stores and therefore fully solving the issues that has
emerged in polyglot persistency environments by joining holistic transactions
with MdsQL. 4 This paper focus on the basic concepts underlying distributed
database system including transaction management in Homogenous Distributed
Real-Time Database System(HDRTDBS). Lastly, this paper helped the organization
when installing homogenous DBMS to implementation of distributed databases and
database software, academicians should also play a major role in exposing
beneficiary to these superior element transaction management. 6 In this paper
the Two-Phase Commit (2PC) protocol for distributed transactions is modeled
with the help of a timed Petri net to analyze the ACID property for consistent
commitment of distributed transactions. After, this research demonstrates a
Petri Net model then, deep-rooted problems are there with the 2PC protocols
such blocking which reduces the availability of the resources. independent
recovery is not possible but the recovery protocols for single site failures.
Ø Types of
Transaction has divided in
contrasting ways. One of them the type of transaction. Transaction classify as
long-life(batch) and short-life (online) 111. The first one online
transaction has a short implementation/time to reply and almost side effects the
small part of the database. The airway booking transactions and Banking
transactions are a good example for this class (online transaction) on the
other hand, batch transaction needed more time for processing and has biggest
part of the database 110. Transaction likewise divided in term of transaction
112. Flat transaction has very important in database where the time of
execution short and a small of the number of concurrent database also the flat
model is very easy and more secure but it has less ability to access the
application require and complex transaction 1. Also, transaction classified
in term of its read-write actions so, the read action executes before write
action so this is called twostep transaction 111. At the same time, called
restricted transaction if the data have been read before have being
written(updated), also called restricted and two-step transaction if a
transaction is restricted and two-step 110.
Ø Transaction Failures
There are some different types of
failure. In the transaction failure can be error when mistaken input
information likewise the detection of a potential deadlock or present 110.
The DMBS must to solve with mixture of type failure including communication failure,
site failure, server failure, and transaction failure 113. Transaction
failure is determined by the Abort. In a local DBMS when the global transaction
aborted, this abort must be propagated to any site then the transaction was
Ø Database Log
The process of updating transaction is
not changing the database but, it is changing and recorded inside the database
Log. Log file also consist of important information to get back the database
and processing can be cause error, subsequently shows to the user. All the
properties are recorded inside the table called log table. Also, the log table
including those attributes 115:
identi?er de?ning primary
key(ID), handled by trigger and sequence.
(event_number) warning or error
(login or ID_program) program
or user identi?er.
of the event itself.
Write-Ahead Logging (WAL)
is an important protocol to controlling logs, if the logs is written
Asynchronously or Synchronously 116. In this case the update on database are
recorded inside the stable memory before log table or log file is altered in a
stable memory to reflect the update 110. If the failure existed before the
log was write so the database will be stay at update form but, the log will not
be shows the update that it is useless to get back the database to a consistent
state and up to date state 110.
Ø Distributed Reliability Protocols
reliability protocol means controlling the durability and atomicity of
distributed transaction when the process on the number of database. Address of the
protocol the distributed execution of the start transaction consists of some
commands such as read, write, commit, abort and recovery commands 117. The
distributed reliability protocol is executed between the participants and the
coordinator. The reliability approach in distributed database management system
(DDBMS) consist of termination, recovery and commit protocols, means that the
implementing of the distributed transaction engages multi- sites, some of them
while the implementing might fail so on the distributed database the effect of
the transaction is one of two nothing or all 110.
Patiño-Martinez, Brondino and Vianello, 2017) they suggest ultra-scalable
transactional management layer when data store with multi-versioning
capabilities with can be integrated. It was integrated with six different data
stores three NoSQL data stores and three SQL-like databases then in the context
of the FP7 CoherentPaaS European Project this layer has been developed 3.
a set of subsystems that they have a play an important role in transactional
processing then consist of three things: common query engine, holistic
transactional manager and data store. The author deals with two points of data
store first to the holistic transactional manager data store fully
representative transactional processing and second data store that make inside
transaction processing and representative of transaction to total transaction
manager for global transactions 3.
The authors (Alkhatib and Labban , 2010)
(Ratawa, singh and Milan, 2017) shows the implementation of Two-Phase commit technique to solving
consistent state problem for the database 45. In the distributed
commit (2PC) protocol is an algorithm to confirm the consistent termination of
a transaction 4. The 2PC protocol have two types of node to finishing the
process like coordinator and subordinate 5. The first phase is a “PREPARE”
phase 45, when the coordinator of the transaction sends the PREPARE
message. The second phase is decision-making phase, when the coordinator
problem a COMMIT message. With one of this method two-phase may be carried out
such as centralized two-phase, linear two-phase and distributed two-phase. The
(Sarkar, Chaki, 2010) present the Petri Net model for two-phase commit protocol
in distribute environment for transaction management. In 2PC have natural
problems like; Blocking. In Time Petri Net model, the time is related to
The time petri
net two periods is used within each transaction. The early-finish-time is
considered to be the minimum and maximum of these times, and late-finish-time
approximately. The triggering interval of resultant transition is the
difference between both Early-Finish-Time EFT and Last-Finish-Time LFT 6.
1 Wang, Ting, Jochem Vonk, Benedikt Kratz, and Paul Grefen. “A survey on
the history of transaction management: from flat to grid
transactions.” Distributed and Parallel Databases 23,
no. 3 (2008): 235-270.
2 Fadia A. Abdalhalim, A. K. TIME, “A SURVEY ON
TRANSACTIONS MANAGEMNET IN DISTRIBUTED REAL TIME DATABASE SYSTEMS.”.
International Journal of Advanced Computer Technology (IJACT) ISSN:2319-7900,
Jiménez Peris, Ricardo, Marta Patiño Martínez, Ivan Brondino, and
Valerio Vianello. “Transaction management across data
stores.” International Journal of High Performance Computing And
Networking (2017): 0-10.
4 Srivastava, Ashish, Udai Shankar, and Sanjay
Kumar Tiwari. “Transaction Management in Homogenous Distributed Real-Time
Replicated Database Systems.” International Journal of Advanced
Research in Computer Science and Software Engineering (2012), P 23-30.
5 Alkhatib, Ghazi, and Ronny S. Labban. “Transaction
management in distributed database systems: the case of oracle’s two-phase
commit.” Journal of Information Systems Education 13, no. 2
Bidyut Biman, and Nabendu Chaki. “Transaction management for distributed
database using petri nets.” International Journal of Computer
Information Systems and Industrial Management Applications (IJCISIM) 2,
no. s 69 (2010): 76.
Waqas, Ahmad, Abdul Waheed Mahessar, Nadeem Mahmood, Zeeshan Bhatti,
Mostafa Karbasi, and Asadullah Shah. “Transaction Management Techniques
and Practices In Current Cloud Computing Environments: A Survey.” International
Journal of Database Management Systems 7, no. 1 (2015): 41.
110 Waqas, Ahmad, Abdul Waheed Mahessar,
Nadeem Mahmood, Zeeshan Bhatti, Mostafa Karbasi, and Asadullah Shah.
“Transaction Management Techniques and Practices In Current Cloud
Computing Environments: A Survey.” International Journal of Database Management Systems 7, no. 1
111 J. Gray and A. Reuter, Transaction Processing:
Concepts and Techniques, Morgan Kaufmann. 1993.
112 P. V. Tamer Ozsu, Principles of Distributed Database
Systems, Springer. 2011.
Breitbart, A. S. (1992). Transaction Management Issues in a Failure-Prone
Multidatabase System Environment . VLDB Joumal, (pp. 1-39).
114 Kumar, Nitesh, Ashish Kumar, and Md Imran Alam.
“Enhanced” One Phase Commit Protocol” In Transaction
Management.” International Journal of Soft Computing and Engineering (IJSCE), (2013) ISSN: 2231-2307.
115 Kvet, Michal,
and Karol Matiaško. “Temporal transaction integrity constraints
management.” Cluster Computing 20, no. 1 (2017): 673-688.
Yacine, and Hamida Seba. “Group key management protocols: A novel
taxonomy.” International journal of information technology 2, no. 1 (2005): 105-118.
117 Özsu, M. Tamer, and
Patrick Valduriez. “Principles of distributed database systems.”