www.ijecs.

in
International Journal Of Engineering And Computer Science ISSN:2319-7242
Volume2 Issue 8 August, 2013 Page No. 2569-2573

Introduction on Data Warehouse with OLTP and OLAP

Arpit Parekh

Maharaja Krishnakumar Sinhji Bhavnagar University, Department of Computer Application, Shree Swaminaranyan Naimisharanaya
College of Management and IT, Bhavnagar, India
aparekh5888@gmail.com

Abstract: Data warehouse, online transaction processing system (OLTP) and on-line analytical processing (OLAP) are basically main
need of the database collection in business, corporate fields and many areas. Nowadays many services, products and new techniques are
available and offering many ideas in the DBMS. This paper is research about the data warehousing with OLAP and OLTP with the
basic need and main parts of the business database management. The main here consider that Data warehouse usage, process, Data
warehouse with Meta data, online transaction processing system (OLTP), on-line analytical processing (OLAP) and also OLTP vs.
OLAP with advantages and disadvantages [5].

Keywords: Meta data, online transaction processing system, on – line analytical processing, Data Warehouse

information directory that includes several enhancing

1. Introduction functions [6].

OLTP, OLAP, Meta data and Data warehouse are essential According to Kelly, a tool for Meta data management should:
elements of supports system, which has increasingly become  Allow administrator to perform system administration
a focus of the database industry. Many commercial products operations, and manage security.
and services are now available, and all of principal database  Allow end users to navigate and query Meta data.
management system providers now offerings in these areas.  Allow end users to uses a GUI.
Decision support places some different requirements on
 Allow end users to extend Meta data.
database technology compared to traditional on – line
 Allow Meta data to be imported / exported into from
transaction processing applications [2].
other standard and formats.

1.3 OLTP: The online operational Database System that

performs online transaction and query processing is called on
– Line transaction Processing (OLTP) systems. Ex. “Day
to day” operations of organizations, such as purchasing,
inventory, manufacturing, banking, payroll registration, and
accounting [5].
Figure 1: The above diagram shows in the integration
system we are use different tools but also database 1.4 OLAP: Data warehouse serve users and knowledge in
management systems need the basic requirements depends the role of data analysis and decision making. Such systems
of Data warehouse, meta data, OLTP and OLAP. can organize & present data in various formats as per needs
of different users. These systems are known as on – line
analytical processing (OLAP) systems [5].
1.2 Metadata

Meta data is “data about data” or Meta data is “the data used 1.5 Data Warehouse: According to Barry Devlin, a single,
to define other data”. It specifies source, values, usage and complete and consistent store of data obtained from a variety
features of DWH data and defines how data can be changed of different sources made available to end users in a way they
and processed at every architecture layer. Meta data is stored can understand and use in a business context. [1]
in a Meta data repository which all the other architecture
components can access. Meta data is a critical need for using,
building, and administering the data warehouse. For end- 2. Research Methodology
users, metadata is like a roadmap to the data warehouse 2.1 Data Warehouse Usage
contents. A Meta data repository is like a general-purpose

Arpit Parekh, IJECS Volume 2 Issue 8 August, 2013 Page No.2569-2573 Page 2569
  Data warehouses and data marts are used in a wide while retaining the rapid implementation and opportunistic
range of applications. application of Bottom up approach.
The design and construction of data warehouse may consist
 Business executives use the data in data warehouses
of the following steps:
and data marts to perform data analysis and make
 Planning
strategic decisions.
 Requirements study
 In many areas, data warehouses are used as an
 Problem Analysis
integral part for enterprise management.
 Warehouse Design
 The data warehouse is mainly used for generating
 Data integration
reports and answering predefined queries.
 Testing and
 It is used to analyze summarized and detailed data,
 Deployment of the data warehouse.
where the results are presented in the form of
reports and charts.
2.3 Software systems can be developed using two
 Later, the data warehouse is used for strategic
methodologies:
purposes, performing multidimensional analysis and
sophisticated operations.
Waterfall method and Spiral method:
 Finally, the data warehouse may be employed for The waterfall method performs a structured and systematic
knowledge discovery and strategic decision making analysis at each step before proceeding to the next, which is
using data mining tools. like a waterfall, falling from one step to next.
 In this context, the tools for data warehousing can he The spiral method involves the rapid generation of
categorized into access and retrieval tools, increasingly functional systems, with short intervals between
database reporting tools, data analysis tools, and successive releases. This is considered a good choice for data
data mining tools [3]. warehouse development, like for data marts, because the
turnaround time is short, modifications can be quickly, and
2.2 Need for Data Warehousing: new designs and technologies can be adapted in a timely
 Industry has huge amount of operational data. manner.
So, the warehouse Design process consists of following
 It is a platform for consolidated historical data for steps:
analysis. 1.] Business process model:
 It stores data of good quality so that knowledge For example, orders, invoices, shipments, inventory, account
worker can make correct / strategic decisions. administration and sales. If the business process us
 Better business intelligence for end-users. organizational and involve multiple complex object
 Reduction in time to locate, access, and analyze collections, a data warehouse model should be followed.
2.] Choose the grain of the business process:
information.
The grain is the fundamental atomic level of data to be
 Consolidation of disparate information sources. represented in the fact table for this process, for example,
 Strategic advantage over competitors. individual truncations, individual daily snapshots, and so on.
 Faster time-to-market for products and services. 3.] Choose the dimensions:
That will apply to each fact table record. Typical dimension
 Replacement of older, less-responsive decision are time, item, customer, supplier, warehouse, transaction
support systems.
type, and status.
 Reduction in demand on IS to generate reports.[3] 4.] Choose the measures:
That will populate each fact table record. Typical measures
2.3 The Process of Data Warehouse Design [5]: are numeric additive quantities like dollars sold and units
A DWH can be built using a Top – Down Approach, Bottom sold [5].
- Up Approach or combination of both.
Top – Down Approach: 2.4 The major feature between OLTP & OLAP [5]:
The top – down approach start with the overall design and 1. User and system orientation:
planning. It is useful in cases where the technology is mature Clerks, clients and information technology
and well known, and the business problems that must be professionals use OLTP systems and it is customer_
solved are clear and well understood. oriented whereas the OLAP is market-oriented used
Bottom – Up Approach: by knowledge workers, analysis and managers.
The bottom up approach starts with experiments and 2. Data contents:
prototypes. This is useful in the early stage of business An OLTP system manages current data and is not
modeling and technology development. It allows used for decision making purposes. An OLAP
organizations to move forward at considerably less expense manages large amounts of historical data with
and to evaluate the benefits of the technology before making facilities for Summerton and aggregation.
significant commitments.
Combined Approach: 3. Database design:
In the combined approach, an organization can exploit the An OLTP system usually adopts an Entity-
planned and strategic nature if the Top – down approach relationship model and an application oriented

Arpit Parekh, IJECS Volume 2 Issue 8 August, 2013 Page No.2569-2573 Page 2570
 database design. An OLAP system uses a star or a  Can leverage functionalities inherent in the
snowflake model. relational database: RDBMS already comes with
4. View: a lot of functionalities. So ROLAP technologies,
An OLTP system works on the current data within (works on top of the RDBMS) can control these
organization, without using historical data in functionalities.
different organizations. An OLAP systems deal with
information that originates from different Disadvantages:
organizations, adding information from many data
stores. Because of their huge volume, OLAP data
 Performance can be slow: Each ROLAP report
is a SQL query (or multiple SQL queries) in the
are stored on multiple storage media.
relational database, the query time can be long if
5. Access Patterns:
the underlying data size is large.
The Access Patterns of an OLTP system consist of
Short, atomic transactions. Such a system requires  Limited by SQL functionalities: ROLAP
concurrency control and recovery mechanisms. technology relies on generating SQL statements
Accesses to OLAP systems are read only operations. to query the relational database, and SQL
2.5 OLAP statements do not fit all needs.
 Logically, OLAP servers present business users  ROLAP technologies are limited by what SQL can
with multidimensional data from DW or data do. ROLAP vendors moderate this risk by
marts, without concerns regarding how or where building the tool with ability to allow users to
the data are stored. define their own functions.
 So an OLAP Server is a high capacity, multi user
data manipulation engine specifically designed 2. Multidimensional OLAP (MOLAP) servers:
to support and operate on multi-dimensional
data structure.
 These servers support multidimensional views of
data through array-based multidimensional
 However, physical architecture and storage engines. They map multidimensional
implementation of OLAP servers must consider views directly to data cube array structures.
data storage issues.
 Advantage of using data cube is that it allows fast
 OLAP stands for On-Line Analytical Processing. indexing to pre-computed summarized data.
 The key feature is "Multidimensional", the ability  With multidimensional data stores, storage
to analyze metrics in different dimensions such utilization may be low if data set is sparse.
as time, geography, gender, product, etc.
 In such cases, sparse matrix compression
 Implementations of a warehouse server for techniques should be explored. Many MOLAP
OLAP processing include the following servers adopt a 2-level storage representation to
 MOLAP (Multidimensional OLAP) and ROLAP handle dense and sparse data sets: denser sub
(Relational OLAP) and Hybrid OLAP cubes are identified and stored as array structures,
(HOLAP) [5]. whereas sparse sub cubes employ compression
2.6 OLAP Types: technology for efficient storage utilization [3].
1. Relational OLAP (ROLAP) servers:
Advantages:
 These are intermediate servers that stand in  Excellent Performance: A MOLAP cube is built
between a relational back-end server and client for fast data retrieval, and is optimal for Slicing
front-end tools. and Dicing operations.
 They use a relational or extended-relational DBMS  Can perform complex calculations: All
to store and manage warehouse data, and OLAP calculations have been pre-generated when the
middleware to support missing pieces. cube is created. Hence, complex calculations are
 ROLAP servers include optimization for each not only feasible, but they return quickly.
DBMS back end, implementation of aggregation
navigation logic, and additional tools and Disadvantages:
services.
 Limited in the amount of data it can handle:
 ROLAP technology tends to have greater Because all calculations are performed when the
scalability than MOLAP technology. cube is built, it is not possible to include a large
 The DSS server of Micro strategy, for example, amount of data in the cube itself.
adopts the ROLAP approach [3].  The data in cube cannot be derived from a large
amount of data. Indeed, this is possible.
Advantages:
 Only summary-level information will be included
 Can handle large amounts of data: The data size in the cube itself.
limitation of ROLAP technology is depends on
data size of the underlying RDBMS. So, ROLAP
 Requires additional investment: Cube
technology is often proprietary and does not
itself places no limitation on data amount.

Arpit Parekh, IJECS Volume 2 Issue 8 August, 2013 Page No.2569-2573 Page 2571
 already exist in the organization. Therefore, to OLAP database there is aggregated, historical data, stored in
adopt MOLAP technology, chances are multi-dimensional schemas (usually star schema).
additional investments in human and capital The following table summarizes the major differences
resources are needed. between OLTP and OLAP system design [7].

OLTP for Operation OLAP for Data

3. Hybrid OLAP (HOLAP) servers: Concepts
system Warehouse
 The hybrid OLAP approach combines Operational data; Consolidation data;
ROLAP and MOLAP technology, OLTPs are the OLAP data comes
benefiting from the greater scalability of Source of data
original source of from the various
ROLAP and the faster computation of the data. OLTP Databases
MOLAP.
 For example, a HOLAP server may allow To control and run To help with planning,
large volumes of detail data to be stored in Purpose of data fundamental problem solving, and
a relational database, while aggregations business tasks decision support
are kept in a separate MOLAP store. The Multi-dimensional
Microsoft SQL Server 2000 supports a Reveals a snapshot
views of various kinds
hybrid OLAP server. What the data of ongoing business
of business activities
 HOLAP technologies attempt to combine the processes
advantages of MOLAP and ROLAP.
Short and fast
 For summary-type information, HOLAP Periodic long-running
Inserts and inserts and updates
leverages cube technology for faster batch jobs refresh the
Updates initiated by end
performance. data
users
 When detail information is needed, HOLAP
can "drill through" from the cube into the Relatively
underlying relational data [3]. standardized and
Often complex queries
Queries simple queries
involving aggregations
3. OLTP vs. OLAP Returning relatively
few records
We can divide IT systems into transactional (OLTP) and Depends on the
analytical (OLAP). In general we can assume that OLTP amount of data
systems provide source data to data warehouses, whereas involved; batch data
OLAP systems help to analyze it. [7] Processing refreshes and complex
Typically very fast
Speed queries may take many
hours; query speed can
be improved by
creating indexes
Larger due to the
existence of
Can be relatively
Space aggregation structures
small if historical
Requirements and history data;
data is archived
requires more indexes
than OLTP
Figure 2: The above diagram shows the OLTP and OALP Typically de-
operations performance and how is difference both with normalized with fewer
Business processes and Business Data Warehouse [7]. Database Highly normalized
tables; use of star
Design with many tables
3.1 OLTP [On-line Transaction Processing] is and/or snowflake
characterized by a large number of short on-line transactions schemas
(INSERT, UPDATE, and DELETE). The main emphasis for Backup religiously;
OLTP systems is put on very fast query processing, Instead of regular
operational data is
maintaining data integrity in multi-access environments and backups, some
critical to run the
an effectiveness measured by number of transactions per environments may
Backup and business, data loss
second. In OLTP database there is detailed and current data, consider simply
Recovery is likely to entail
and schema used to store transactional databases is the entity reloading the OLTP
significant monetary
model (usually 3NF). data as a recovery
loss and legal
method
liability
3.2 OLAP [On-line Analytical Processing] is characterized
by relatively low volume of transactions. Queries are often
very complex and involve aggregations. For OLAP systems a 4. CONCLUSION
response time is an effectiveness measure. OLAP In Database management system multiple requirements are
applications are widely used by Data Mining techniques. In available and include new business techniques.

Arpit Parekh, IJECS Volume 2 Issue 8 August, 2013 Page No.2569-2573 Page 2572
In above discussion we see that the Business Database
management and Data warehousing depended on the
performance of Meta data, OLTP and OLAP performance.
In the data warehousing move toward, information is
requested, processed, and merged continuously, so the
information is readily available for direct querying OLAP
and analysis at the warehouse.

5. REFERENCES
BOOKS:
[1] "The Story So Far". 2002-04-15. Retrieved 2008-09-21.

[2] Data Warehousing in the Real World;

Sam Anahory & Dennis Murray; 1997, Pearson

[3] Data Warehousing, Data Mining and OLTP;

Alex Berson, Stephen J. Smith 1997, McGraw Hill
Page: 4, 14-16.

[4] Data warehousing System; Mallach; 2000, McGraw

[5] Data Mining – Concepts & Techniques; Jiawei Han &

Micheline Kamber, Jain Pei, Morgan Kaufmann publication;
Third Edition; Page: 125 to 130, 151,152, 164, 165

[6] Metadata Standards and Metadata Registries: An

Overview" (PDF). Retrieved 2011-12-23.

[7] http://datawarehouse4u.info/OLTP-vs-OLAP.html

[8] An Introduction to Database Systems; C. J. Date,

A. Kannnan, S. Swaynathan

WEBSITES:
[8] http://en.wikipedia.org.

[9] Google Scholar. http://scholar.google.com/scholar.

[10] My blog: www.aparekh5888.blogspot.com

Author Profile

Parekh Arpit received the Msc.it and PGDCA degrees in

Information Technology from Bhavnagar University and
Computer Application from Saurashtra University in 2011
and 2009, respectively. Works at shree swaminaranya
Naimisharanaya College of Management and IT.

Arpit Parekh, IJECS Volume 2 Issue 8 August, 2013 Page No.2569-2573 Page 2573