Transaction Processing Systems

by Julian Carpo

Transaction processing systems

By Julian A. Carpo

by Julian Carpo

Characteristics of transaction processing systems

Batch transaction processing Real-time transaction processing Data validation Historical significance of transaction

processing systems Manual transaction systems

by Julian Carpo

Characteristics of transaction processing systems Transaction: event generating or modifying stored data

in an info sys. Batch transaction processing collects transaction data

as a group/batch processed later (time delay) Real-time transaction processing: immediate data

processing Data Validation: used to check entry of transaction data Historical significance of transaction processing

systems: UNIVAC Manual transaction systems: business systems that

operate without the use of machines.

by Julian Carpo

TPS Characteristics

They collect, store, modify and retrieve the transaction of an organisation.

Directly support business operations Must be designed in conjunction w/

the organisation’s procedures. Main information processes: collecting

and storage.

by Julian Carpo

4 important characteristics Rapid response (fast performance, rapid

response time) Reliability (low failure rate, quick and

accurate recovery, backup and recovery procedures)

Inflexibility (every transaction processed in same way regardless of conditions)

Controlled processing (supports an organisation’s operations)

by Julian Carpo

Examples of batch transaction processing

Clearance of cheques Other forms of paper output

by Julian Carpo

Large batch

Mainframe PC and magnetic tape Payroll or stock info Employee: hours worked, overtime

earned Payroll master file updated Pay slips created (ALL employees) Often run at night – less demand for

info sys

by Julian Carpo

3 disadvantages in batch processing

Processing schedule predetermined Errors unable to be corrected during

processing Sorting transaction data – expensive

and time consuming

by Julian Carpo

2 concerns w/ real-time transaction processing

Concurrency: data cannot be changed by 2 users at same time

Atomicity: steps completed succesfully as a group. If any step fails, no other step should be completed.

by Julian Carpo

Transaction processing monitor

Software (middleware) allowing running of TP programs.

Manages sequence of events Provides interface between input

devices and DBMS Provides data security, ensuring

transactions don’t get lost/corrupted.

by Julian Carpo

Differences btween real-time and batch Real time: each transaction is unique; transactions are

stand-alone; requires master file to be available more often for updating; fewer errors – transaction data is validated and entered immediately; infrequent errors may occur – but often tolerated; not practical to shut down whole system (infrequent errors); more computer operators are required in real-time processing – operations not centralised.

Batch: each transaction part of a group; database not accesible all of the time; more errors; data is organised and stored before master file is updated – errors can occur during these steps; easier to maintain than real-time.

by Julian Carpo

Data validation

Used to check the entry of transaction data.

Procedures ensure transactions are correct and have been accurately stored in the database.

Involves transaction initiation and field checking.

by Julian Carpo

Historical significance and manual transaction systems

Historical significance: UNIVAC 1950’s Manual transaction systems: business

systems operating w/o use of machines.

by Julian Carpo

Components of a transaction processing system

Users: take data and use it in another info system type

Participants: conduct info processing People (from the environment):

directly enter transactions and perform validation

by Julian Carpo

Examples of real-time transaction processing

Reservation systems: set aside service/product for future use

Motels Point-of-sale (POS) terminals: sells

goods/services Library loan systems: keeps track of

items borrowed from library

by Julian Carpo

System flowchart for a POS system

Server POS system

Customer receipt

Product Database

UPC Price

UPC/quantity

UPC/quantity

Inventory database

Purchase database

MIS

Inventory report

by Julian Carpo

Examples of batch transaction processing

Cheque clearance: written order asking bank to pay money to a person

Bill generation: an invoice for goods/services supplied to a customer.

Credit card sales transaction: takes impression of customer’s card

by Julian Carpo

Databases and files

Database: collection of data Types: hierarchical, network,

relational Important features for design: good

data placement, short transactions, real-time backup, high normalisation, archiving of historical data and good hardware configuration

by Julian Carpo

File: block of data Master file: info about business

organisation Transaction file: collection of transaction

records Report file: data formatted for presentation Work file: temporary file in the system used

during processing Program file: instructions for processing of

data

by Julian Carpo

Data warehousing

Data warehouse: collects info from different data sources

Consolidated, subject-oriented, historical and read-only

by Julian Carpo

Backup procedures

Backup: another copy of data Recovery processes include backup,

journal, checkpoint and recovery manager

by Julian Carpo

2 types of recovery

Backward recovery: back out or undo unwanted changes

Forward recovery: redo changes

by Julian Carpo

Recovery processes continued

Magnetic tape: stores large data quantities inexpensively. Uses sequential access

Grandfather-father-son: backup procedure refers to at least 3 generations of backup master files

Partial backups: parts of master file backed up

by Julian Carpo

Updating in a batch Originally feasible: details were stored on

punch cards or magnetic tape 2 stages – collecting and processing of

transaction data into file & updating master file

Sequential access: data accessed in a sequence. Time consuming.

Information technology: magnetic tape

by Julian Carpo

Updating in real time Direct/random access: data is accessed

without accessing previous data items. Stores data based on an algorithm, which calculates data location. When index used, called indexed access.

Information technology: magnetic disk Software – online and user-friendly Rapid response time – few seconds or less

by Julian Carpo

Other information processes

Collecting: generating transaction data

Analysing data: meet user’s information needs. Output from TPS is input to other systems

by Julian Carpo

Hardware

MICR (Magnetic ink character recognition): used by banks to read account numbers on cheques

ATM (Automatic teller machine): banking terminal performing deposits and withdrawals

Barcode readers: used in retail to collect product information

by Julian Carpo

Forms

Document used to collect data from a person

Examples include sign-on sheet for payroll.

On-screen forms: populates fields in a database

Web forms: used to purchase items over the internet

by Julian Carpo

Decision support systems

Assist people to make decisions by providing information, models and analysis tools

Data mining: used in DSSs to find relationships and patterns in the data

by Julian Carpo

Management information systems

Provides information for the manager Different types of reports include

scheduled reports, forecasting reports, on-demand reports and exception reports

by Julian Carpo

Key issues in designing an MIS or DSS

How a system will be modeled How model of system will be handled

by the computer What data will be used How far into the future trends will be

extrapolated

by Julian Carpo

Issues related to transaction processing systems

Nature of work Non-computer procedures Bias Importance of data Control in transaction processing

by Julian Carpo

Nature of work

Automation of jobs: the use of information technology to perform tasks once performed by people

People as participants: resulted in fewer jobs in banks. People are purchasing products by completing a web form.

by Julian Carpo

Non-computer procedures

Organisations rely on TPS Manual backup procedure if system

crashes When system up & running, user

needs procedure to enter manual transactions

by Julian Carpo

Bias

Data unfairly skewed or gives too much weight on particular result

Data gathered can be presented in biased way using tables and charts

Becomes ethical issue when relevant information is misrepresented

by Julian Carpo

Importance of data

Data security: involves series of safeguards to protect data

Encryption: coding data Decryption: changing it back Firewalls: used on networks to verify

and authenticate all incoming data

by Julian Carpo

Importance of data continued Accuracy: extent data is free from errors.

Errors caused by mistakes in gathering data, data entry, mismatch of data and person, or out-of-date information

Data validation: checks data entry. Checks for missing data, too high or too low data values, data values inconsistent with other data, data in wrong format.

Carried out using range checks, list checks, type checks and check digits (see Chapter 2)

by Julian Carpo

Data integrity

Describes reliability of data. Involves accuracy, currency and

relevance of data

by Julian Carpo

ACID test Atomicity: all steps involved in transaction

completed successfully as a group Consistency: transaction successfully

transforms the system and database from one valid state to another

Isolation: transaction is processed concurrently with other transactions. As if only transaction executing the system

Durability: all changes transaction makes to database become permanent

by Julian Carpo

Control in transaction processing

Starts with collecting, includes manner TPS manipulates data & way errors are corrected

Some people in management positions are willing to falsify transactions to promote careers

by Julian Carpo

Aspects of TPS relevant to “A large retail company” Real-time transaction processing Manual transaction systems Components of a TPS Credit card sales transactions Database and files Updating in a batch, real-time Other information processes: collecting Nature of work Non-computer procedures Bias

by Julian Carpo

Real-time transaction processing It is important in an organisation such as a “large

retail company” for the immediate processing of data & instant confirmation of a transaction (specifically needed in a retail environment where servicing a number of customers with a limited staff under a limited amount of time is usual). The 4 important characteristics of a TPS stresses this. They are rapid response, reliability, inflexibility and controlled processing. Also, the concerns of concurrency and atomicity are important in ensuring data is not mishandled or any procedures/steps taken are completed successfully as a group or not executed at all.

by Julian Carpo

Manual transaction systems Manual transaction systems are imperative in the retail environment

because they were the conceptual basis for real-time TPS. They were used long before the rise in technology allowed for easier data collection and storage. Obviously, they have been around since trading and business took its early strides. In smaller stalls who still cannot afford the technology corporations use, it is necessary to use manual transaction systems to record business activities. Manual transaction systems are based on a clear set of rules followed by a person. They make it easier to transmit the data into a format suitable for recording onto an information technology e.g. a computer. Of course, computerisation of a manual T.S. brings benefits, when used by a large retail organisation. It increases the rate at which products are sold, which means more profit for the retail chain. The less time taken for a customer to purchase a product is invaluable in customer satisfaction, as there are lesser queues – therefore lesser time spent on waiting. Also manual T.S.s provides the retail chain with information in what products are on demand. The retail chain capitalises on this information on consumer habits and they can tailor their products to attract more customers, thereby meaning more profit.

by Julian Carpo

Components of a TPS People are obviously important in an operation of a TPS.

They are usually classified as users, participants and people in the environment. This classification allows for easier role recognition within the retail chain organisation. Users do not interact w/ the TPS but use the data provided. They provide information about the system that is useful to the large retail chain. The participants are the shop workers which do the actual work conducted in a retail chain store. They enter the data into the TPSs. And the peole from the environment are the consumers, or people who buy goods from the shop front. They become participants when they use EFTPOS terminals located within the retail store to purchase goods and ATMS within or outside of it. They become participants in the sense that they directly enter transactions and perform validations.

by Julian Carpo

Examples of real-time transaction processing

Reservation systems basically involve setting aside a product or service for future purchase or use. They are common in the retail industry. Lay-by is a good example. This involves the delaying of a valued item for future purchase. A minor monetary deposit, usually a small percentage of the overall purchase price of the product is then used to secure the customer the right to purchase this product at a later time.

by Julian Carpo

Examples of real-time transaction processing continued

POS terminals are used by retail stores to sell goods and services. In large retail organisations, POS terminals send inventory data to a central computer (sale made). We as users become participants when we use these machines to purchase items from the store. The data is converted for easier transmission. Relevant information and the price are displayed using barcode readers.

by Julian Carpo

Examples of real-time transaction processing continued

Credit card sales transactions can be done over the internet, at POS terminals or taking an impression of customer’s credit card. Both the POS system and the multi-page credit slip system rely on batch processing. We use them as an alternative to money, to pay for retail store purchases.

by Julian Carpo

Databases and files

In any large retail chain/organisation, the information processes often vary. In the handling of data, the storage and retrieval of data must be spot on. Retail chains rely on this trait for the success of their business.

by Julian Carpo

Databases and files

Good data placement:Database is designed to efficiently gain access to frequently used data & data patterns

Short transactions: transactions can be processed fast

Real-time backup: saving data during low usage or idle time

by Julian Carpo

Databases and files continued High normalisation: data redundancy is

minimised to increase update speed. Also improves backup speed.

Archiving of historical data: old data stored in chronological order in separate databases.

Good hardware configuration: the setup must be able to handle a number of users and provide rapid response

by Julian Carpo

Files Master file: info about a retail chain’s

business. This is where the data is stored Transaction file: audit trails and history for

the retail chain Report file: formatted user presentable data Work file: temporary file Program file: data processing instructions.

by Julian Carpo

Backup procedures

Backups provide safety nets for data restoration in case of a system crash/malfunction. Without these, the retail chain will cease to make profits as a result of data loss. They are usually stored in a safe location.

by Julian Carpo

Recovery process When a TPS fails, data must be recovered

immediately for restoration. Journal: keeps track of company records on

a physical medium. Checkpoint: are like signposts. Indicate

where, when, what data was saved in a daily period. Syncs files and journals together for smoother transaction reviewing

by Julian Carpo

Recovery process

Backward recovery: undo database changes

Forward recovery: redo database changes

Recovery manager: restores database to a previous condition

by Julian Carpo

Magnetic tape

Often used. Inexpensive. Appropriate for storing large amounts of data. An alternative to hard disk drives.

Data is read from the beginning until the end.

Suitable for most transaction backups

by Julian Carpo

Grandfather-father-son

3 generations backup procedure. Son is the most recent, whilst father and grandfather are the oldest. Commonly used in conjunction w/ magnetic tape. Ensures data cannot be lost. A large retail chain should have more than 3 generations.

by Julian Carpo

Other information processes

Collecting: entering the transaction data. The users who become participants by using EFTPOS or ATM are the data source

MICR readers are used by banks to read account numbers on cheques

ATM: allows deposits and withdrawals for customers

by Julian Carpo

Other information processes continued Barcode readers collect product information. They

scan the barcode on the product , the details kept in a central computer. Data about the item is quickly collected.

Forms: documents to collect data from customers. We use this to create memberships, gain customer details and update information. There are paper forms, such as sign on sheets; on-screen forms for computerised population of fields in a database; web forms used by consumers to buy items over the internet.

by Julian Carpo

A large retail company needs a project plan to effectively organise the components of this project. We need a schedule for both the project and subprojects. It is important because it will help to deliver this new solution within feasible requirements and will minimise the effect of unanticipated problems. We need a project goal, a deliverable and a schedule to foresee how this implementation will pan out. A large retail chain cannot afford to compensate for costly mistakes produced by inorganisation and undelivered goals. This will cost us in profits and on a greater extent, the functioning of our business.

Information management software will help to manage information and schedule tasks. Appointments and meeting within co-workers and the organisation can be viewed electronically.

Gannt charts will be used, as they provide users a quick method of determining whether the project is on schedule.

Planning the new solution

by Julian Carpo

Designing the new solution A top-down approach is best to designing this new solution,

as problems that may be encountered along the way can be divided into smaller segments to make solving the problem easier.

We will ensist advice and support for programming this new solution from the information systems personell. We will then adapt a uniform structure for the real-time systems that will be used in the retail stores - based on a popular application - to suit our needs. Reliability is crucial, as this new solution will handle numerous transactions for the working/non-working weeks. Over time, we will train our applicants/employees/participants in handling this new solution. As such, the UMDIT stages in the development cycle will be used.

by Julian Carpo

Designing the new solution continued Context diagrams provide an easy way to visualise

information. Consulting this design tool will help in simplifying the development of the new system, as it clearly shows the flow of data in the system.

A DFD will be used to provide more detail into what the participants can do in regards to this new solution. It has greater detail than the context diagram.

Decision trees are very important in this retail context. Almost anyone can immediately identify what will transpire with this solution, what needs to be done in order to accomplish the goals, etc. it represents all possible decisions and the outcomes of those decisions.

A Data dictionary provides a comprehensive list of the data.

by Julian Carpo

Implementing the new solution Participant training, as outlined before, will come in handy

in ensuring that our employees and workers can use the system and understand its benefits. Training manuals will be accessible. Technical support in troubleshooting problems will be gained online through an always-on internet connection.

Pilot conversion ensures that all our retail chain stores across Australia will have the latest technology available to them concurrently. This also ensures that the system can be trialled in different regions e.g. rural as opposed to urban. If successful, the new solution is guaranteed to be functioning within acceptable standards and will be implemented throughout the larger centres.

Of course, testing the new system ensures that it works. Diagnostic software and general operation are examples of testing. The reliability of the backup system is integral in the new system’s success.

by Julian Carpo

Bibliography

Heinemann HSC Information Processes and Technology - G.K. Powers