CHAPTER 1 - SYSTEM FUNDAMENTALS Flashcards
System life cycle
The stages through which the development of a new system passes through
5 stages of system life cycle
Analysis, Design, Implementation, Operation, Maintenance
Why is new system created
To replace inefficient old system, fix default/defects, increase quality of output/productivity of the system
Requirements needed for a new system
Internet, RAM, operating system, hardware, software
Limitation of new system
- end users not being able to use the system properly
- data migration is time consuming
- memory issues and data clashes when system is developed/tested
- maintenance is costly overtime
Potential organizational issues in planning of a new system
- lack of attention to required training
- lack of attention to organizational issues
- lack of stakeholder and end-user participation
Whats a feasibility study
study that evaluates and analyzes a project and its potential based on TELOS criteria: Technical, Economical, Legal, Operation and Scheduling
Planning
process of thinking about various details and organizing the activities required to achieve the desired goal
Whats TELOS criteria
Technical: Is existing technology appropriate to implement proposed system
Economical: Is it cost-effective
Legal: Is it against any laws/regulations
Operational: Is existing organizational practices sufficient to maintain and operate system
Scheduling: How long will we wait
What is change management
Involves various parameters and is the process of shifting existing individuals, teams and departments from an existing to desired state; goal is to maximise benefits and minimize negative impacts of the change on individuals
What is a legacy system
Old technology, hardware or computer system
Why are legacy systems still used
- Data cannot be converted to newer formats/applications cannot be upgraded
- System provides users needs even if newer systems exists
Outline one problem of maintaining legacy systems
Continuous patching, updating and upgrading of the old system to look out for any bugs/errors
4 strategies to integration
- Replace both information systems with a new one (high initial cost)
- Keep both information systems and develop them to have the same functionality (high maintenance cost)
- Select one company’s computer system and discard the other (policy problems)
- Select best info system from each company and combine them (workers may have difficulties working with a system from another company)
Software incompatibility
a situation where 2 different software systems cannot operate cooperatively/individually in a same computer or different computer linked by a LAN/WAN (network)
SaaS “Software as a Service”
a contemporary delivery method that allows software and data to be hosted & managed centrally on a remote datacenter.
Example: Powerschool/Netflix; basically just a software that offers a service
Features of SaaS
- Subscription: users have to pay to access services provided
- Uses a multi-tenancy model: single software and its supporting infrastructure can serve multiple customers
- Uses cloud service to store customer data (which reduces data loss risks)
Limitations to SaaS
- Possible data loss if SaaS provider goes out of business
- Involvement of a third party i.e. SaaS provider which risks user data privacy
- Uncertainty of data’s location: customers have lower access to their personal data; usually stored away in datacenter
- Risk of identity theft: the providers often require credit card payments for subscription; may not provide better identity management security/poor firewall
- Connectivity requirement: SaaS model is based on web delivery, if internet service fails, loses access to software data
- Access management: if design of the access control system is inefficient, a single act of accessing the cloud service can expose confidential data
Distinguish between Phased and Direct changeover
Phased conversion method is when the new system is implemented/introduced per module to the whole company. If the particular program/module is proven satisfactory, another module will be introduced until the system is introduced completely. With this approach, the training period is extended and the adoption of the new system takes longer since it is implemented separately at different times. On the other hand, direct changeover is very risky since the company replaces their old system with a new one immediately. With this approach, all users require proper training before the switch takes place, in order to use the new information system efficiently.
Direct/Big Bang changeover
instant from old to new
Parallel changeover
old and new is run side by side; only when output of new is proven satisfactory then the old system is terminated
Pilot changeover
the whole system is introduced to one branch/department of the company and its performance accessed before being introduced elsewhere in the company
Phased changeover
the new system is introduced per module to the whole company, only when its proves satisfactory then the next part is introduced - until the old system is fully replaced
Data migration
transfer data between different formats, storage types, computer systems; It usually takes place when a business changes, upgrades, or merges its computer systems
Possible problems during data migration
- Data loss: due to incomplete data transfers/ errors because of differences in International conventions concerning data time and measurement units
- Data incompatibility: information during data migration may not retransferred due to parameters such as incompatibility with the new systems/non recognizable data structures
2 types of main testing
Functional testing = test whether or not the features of the program (menu functions, individual commands, text input) perform according to design specifications. For example, the programmer wishes to test if the system’s function “add to client” brings them to the add client module
Data testing = involves using normal, abnormal and extreme data to test the system. For example, normal data “75” is used to test whether or not the output of the school’s marking system (with 60 as the middle range) will show “pass” or fail” messages appropriately
Alpha testing
- Testing the software before the product s released to public
- Tests by the company in lab type environment and not by end users in their usual workplaces
Beta testing
- include comments and suggestions from end users
- their feedback valuable to fix defects/errors of the program & contribute to the preparation of support teams who will deal with expected problem
Dry run test
conducted using pen-paper by programmer; run the algorithm mentally and decides on what the output of a run should be (execution)
Unit test
individual parts of the system are tested seperately
Integration test
entire system is tested at the same time to verify that all components work together
User acceptance test
determine if system meets user needs and conducted within the users premises before accepting transfer of ownership
Debugging
process of identifying and correcting errors/faults in a new system