How System Architects cover the whole

The success of a project is inseparable from people with clear division of labor and hard work. However, it is undeniable that one person must cover the whole site, so as to comprehensively think about and deploy the entire site, evaluate and identify various potential risks and adopt further strategies and measures. Is this person you?

First of all, does this person have a key point for you to position yourself? If you only want to cover the part you are responsible for, then your role can only be limited to one, if you have great passion and motivation to cover the whole, then this role may be you.

So from top to bottom, what are the major aspects that a site or project needs to focus on?

 

1. user needs, mining user needs and market needs, and essentially grasping business processes. It is even better to grasp the essence of an industry.

Do you need to determine if this product feature is really needed by the user? Is it just the subjective judgment of the product manager? Have you done any research? Is this feature promising?

 

2. user experience. This is a bit similar to user requirements, but the coverage is more detailed, such as the page opening speed, the interface is friendly, the user needs to get timely information and prompt whether it can be obtained in a timely manner, do you need to click once more? How many steps do you need to jump to the core process? Can it be simpler? Is user registration a high threshold? Do I need to fill in more information? Do all of these information have to be entered at one time? Can I delay the input so that users can register quickly first? Is logon Threshold High? Do users frequently enter verification codes? Are there too many page texts and different page styles? Can the help information help users? Are there any typos? Is the link 404? Is the page error prompt friendly?

 

3. Security requirements such as system crash prevention.

What are the potential security vulnerabilities in the system, XSS? Is there a possibility of being destroyed? Is there a limit on the frequency of key points? Is the Key page automatically monitored? System Process and port monitoring, system traffic monitoring.

 

4. Daily operation.

Operation is the lifeline of Internet products. Several considerations should be taken into account for operation:

(1) operating costs. Hard operating costs, such as machine costs, bandwidth costs, and power consumption costs; soft operating costs: whether to automate the operation and provide good operating support tools and systems to provide operational efficiency;

(2) daily operation specifications and Process Construction: how to implement the process of mounting an item? What about dismounting a merchant? What about proportional adjustment?

(3) operation scale-up: Can I scale up easily according to set?

 

5. Fault Tolerance and Disaster Tolerance considerations.

Is the system flexible and available? With the deterioration of the external environment, can the system provide grayscale services automatically? Does the implementation of system features take into account the insertion of multiple levels and the setting of switches at various levels, depending on the external situation, the relevant functional modules can be phased out? Has the announcement and user prediction function been implemented? Is cross-IDC deployment achieved? Is there a single point of failure? Is static and dynamic separation achieved? Is interaction and presentation separated?

 

6. Data collection.

Internet products must collect various types of data, such

(1) user behavior-related data, such as PV, UV, hotspot click rate, and redirection flow.

(2) website performance data: for example, speed reporting of page opening, automated testing of CGI response time, etc.

(3) Report various security monitoring data