Basic modeling knowledge study Note 3

Physical Design: analyze the details and plan how to implement the solution. The output of the logical design step is the input of the physical design step.
The logical design steps are used to obtain the business object model and a list of business objects and services. The output accurately describes the implementation content, but uses the logic method. Now we need to convert these models and descriptions into the diagrams and descriptions required by developers to implement the solution. In other words, the diagram and description must be detailed enough to forward to developers and enable them to start implementing the solution. Further refine the physical design and improve the technology until the project team feels that developers can implement the solution based on this physical design.

Steps required for Physical Design:
1. Select Technology
2. Build a physical database model
3. Build a physical. net service model
4. Optimize the system process
5. Deploy services

Select the constraint category for the technology:
1. usage requirements:
User features
User skills
2. Scalability:
How many concurrent users are supported?
What is the total number of users?
What is the impact on the future?
3. maintainability
Identifies the number and frequency of future changes.
Overall Maintainability (including update and Error Correction)
4. Reliability:
How high is the possibility?
Can I accept system crashes?
If yes, how high is the acceptance level?
5. performance:
Response time
Overall Performance
6. Security:
What is the confidentiality of system data?
Is it about data encryption?
7. developer knowledge and experience:
8. Future Planning
How will the company's solutions and institutions develop in the future?

Users' expectations for reliability, speed, and security are often recorded in Service Level Agreement (SLA.

Maintainability (how easy it is to modify and continue to work)
Scalability (how simple is the new feature)

A large number of questions are required:
How many users are there?
How many concurrent users are there?
How much data storage is required?
What are the requirements for response time?
How high is the level of data confidentiality required?
What is the availability level of the required solution?

The data storage part should be designed as early as the physical design starts.

Three-tier service design:
1. Data Service. Its task is to retrieve data from the database and send the data to the caller.
2. Business Service: Call data from data service. It does not make any changes to the data. It is transmitted only.
3. User Service: receives and displays data to users.

Interfaces are critical to services. The service must understand how to communicate with other services, and the communication task is completed by the interface.

Design Optimization:
1. Test user scenarios
2. Prototype
3. Analyze the commonalities of development blocks
4. Observe the audience

Planning solution:
1. Schedule: To determine the start date, end date, and related resources, a project budget should be prepared, not only development resources, but also all costs of the project, and license, hardware investment, software investment and training. The general schedule should be expressed by hour and workday.
2. milestone: "Milestone" indicates the finished work and is used as the intermediate state point of the project process.
3. Communication Planning: Generally, the customer can only communicate with the product manager (customer project manager. Without the participation of the product manager, the customer is not allowed to establish the Protocol directly by accessing the developer. It is also necessary to determine which form (written or verbal) of the change request is acceptable. Another important question is determining qualifications. Who can sign the project documentation? Is it the customer or the product manager specified by the customer?
4. iterative planning: enterprise software development needs to be implemented in iterations to maintain focus and simplify and accelerate the process. In the early stages, You need to determine the iterations, the resources involved, and the schedules involved. An important rule for building iterative planning is that the first iteration should contain the required core functions, and the second iteration should contain the most important features (starting from the characteristic with the highest risk ). In this way, the overall project risk will be reduced during the project process.
5. Test Plan: Determine the start and end dates of the test and how to report and resolve the problem. It is necessary to notify customers of their participation methods, including dates and required work. Determine the impact of failure to meet these requirements.

Important: The Trade-off Strategy may ignore the later features. Therefore, when building iterative planning, we should start with the core features. Among these core features, the development priority should be determined based on the risk arrangement.

Record Solutions should begin early, and should begin to record the solutions in the planning phase during design: record the use cases and scenarios. Documented documents must be archived electronically and accessible.