A dozen dormitory buildings, each hundreds of people, and a system for the WiFi network fees in a boarding house, can drag down the network centre if one is deployed independently. We have seen the most typical confusion: A system with a manufacturer's first, B with a partner who doesn't access the Internet, financials two statements, and two backstage codes to the carrier. So central control is not an option but rather an option as long as the dormitory exceeds three or five. The focus here is not on stacking equipment on a machine, but rather on managing flat alignment, data harmonization and strategic harmonization.
Management flat alignment is the first principle.
The most tangible benefit of centralized control is that there is only one back office, a system of accounts and a billing summary. Students use the same account, the same set of meals, and the same balance, regardless of the building they live in. Network centre teachers can see online numbers, traffic distributions, arrears per building at one interface, without going to the next. When we do this, all buildings have access controls registered under the same billing platform, and the building is just a logical partition, not a physical silo.
The strategy is to be under the building.
The situation is very far away from the building: the post-graduate buildings may allow greater traffic, visitors' buildings need temporary accounts, and registrations should be opened for two weeks at the start of a new building. Good central control is that the strategy centre is designed in a coherent manner and the building labels are flexible. We generally recommend that the building be labelled as a type of building, such as a new building, a senior year building, a visitor complex, with different billing rules and speed limits.
Data aggregates to be real or near.
Bills and traffic data are collected, many schools are tangled in real time or in real time. It is cool when it is true, but every flow bill returns to the centre, putting pressure on the network and database, and late peaks are easy to block. We recommend that it be done in real time: the control device at the building side is locally charged, pre-controlled, back-to-back with aggregate data every few minutes, and the center is responsible for general check-out and reconciliation. This ensures both the timeliness of network breakout controls and the central side data volume is manageable.
The fault quarantine is more important than concentration.
Centralization has a hidden risk: the center is hanging, the school is paralysed. So the architecture must be built into it. We usually ask that each building's access controller retain its basic billing capacity locally, that the centre is cut off and that the building side continues to charge and control data for local buffer strategies, but not transmit data at short notice. The centre resumes retransmission. This design is conservative, but saved many times in actual operation, with one central database upgrade and more than 10 buildings where students are connected to the Internet without detection.
You have to be seamless in your travels.
Students cannot just stay in their own building. They can access the Internet by going to a dorm room, to a study room, and to a cafeteria. Under a centralized system, students enter the same number of the certified pages of any building, which they have to register for, and both the certification and the billing costs do not need to be re-registered. We do this with a special look at cross-building: from Building A to Building B, automatic reconnection of equipment, automatic entry into account, uninterrupted billing. This is a bad move, and students complain that the building has been switched off and the passenger service is on its way. Half of the value of central control is reflected in this insensitive travel.
The power is graded to different characters.
The backstage of the building may need to see the building’s online numbers and press entrances, the network centre teachers look at the whole story, the financial statements summary but not the equipment. We suggest that we limit the role before going on line and avoid anyone holding full authority and blame for what happened. The level of authority is not complicated, but it reduces the amount of mishandling and responsibility.
It's smooth.
The dormitory will be built, and the central control platform will be smooth. We suggest that the building should be uncoupled by a structure that would allow access controllers and central platforms to be added to the centre, which is not to touch the center core. The center side will have less performance than the horizontal nodes, without having to push it backwards.
We need all audit logs.
Under central control, who changes the strategy of which building, who manually adjusts a student’s bill, must leave a mark. First, the problem is traceable and the internal mishandling or irregularities are not documented. We will use the operation audit as a mandatory requirement, with logs kept for at least one year. The concentration brings ease and risk of centralized access, and the audit log is that insurance policy.