What is the general functional attribute design of financial data center network (1)

Fancy Wang 0426 2022

Basic information design refers to the design of basic attributes of network devices, including IP address planning and design, VNI and VPC planning and design, and device naming conventions.

IP address planning and design

The planning principles of IP addresses need to consider standardization, scalability, continuity, flexibility, security, and schematicity, and comprehensively consider factors such as business development and IP address allocation status, so as to minimize the complexity of adjusting IP addresses in the future.

Normative: The spatial division and allocation of network addresses is based on a comprehensive understanding of applications, data flow, and user access. It guides and enforces planning, allocation, and use of IP addresses. IP address resources should be managed and allocated in a unified manner across the entire network.

Scalability: The network address has the characteristics of easy expansion in terms of function, capacity, coverage capability, etc. to meet the requirements of the rapid business development on the infrastructure, and the allocation of IP addresses at each level must leave a margin , the continuity required for address aggregation can be guaranteed only when the network scale is expanded.

Continuity: Continuity of addresses required for address summarization is guaranteed as the network scales. The network address space division adopts a hierarchical and hierarchical structured method. For network components or service types with the same function, try to allocate continuous IP address space to facilitate route aggregation and security control, simplify routing tables, and improve the efficiency of routing algorithms. .

Flexibility: IP address allocation should be flexible to meet the optimization of various traffic policies, security policies, routing policies, and make full use of address space.

Security: Internal addresses must not be leaked on the public network, external addresses and internal addresses at the network boundary use different network segments, and NAT functions are deployed as needed.

Indicative: IP addresses reflect the roles, locations or business attributes of network devices or servers (terminals) as much as possible to facilitate identification and management.

IP address allocation principles include the following two aspects:

• The variable-length network number and host space address mask are used to allocate according to the number of hosts in the network segment plus a certain amount of reservation to ensure address aggregation and reduce unnecessary waste of IP address space.
• IP address allocation should facilitate route summarization. Therefore, devices in the same network area should be assigned to the same IP address segment as much as possible. If this is not possible, it is necessary to ensure that network routing can be summarized at the network aggregation layer.

For different types of addresses, IP address classification and allocation methods are also different.

Loopback address: The loopback address is the IP address of the loopback interface used for network management. To facilitate management, you need to create a Loopback interface for each Layer 3 network device, and specify an IP address on the interface as the management address. The network device of the US network unit (service center, control center, etc.) uses the smallest IP address in the address segment space divided for the network device as the loopback address. The loopback address uses a 32-bit masked address. The more the core device is, the smaller the loopback address number should be.

Interconnection address: The interconnection address refers to the address required by the interface where two or more network devices are connected to each other. The addresses assigned to devices at different levels should be different. Relative to the core devices, addresses with smaller numbers are used. The interconnection addresses are usually aggregated and then released. When planning, full consideration should be given to the use of consecutive aggregated addresses. For interconnected devices with point-to-point interfaces, use a 30- or 31-bit masked interconnect address (2 available addresses). For interconnected devices with point-to-multipoint or broadcast type interfaces, use a 28-bit masked interconnect address (14 available addresses).

Service address: refers to various servers connected to the Ethernet, the address used by the host and the address of the gateway. The gateway address of the business address uses the same last digit (such as 254, 126, etc.). If VRRP needs to be deployed, reduce the last digit by 1 (such as 253, 125, etc.) compared with the address used by the core or main equipment. Devices use addresses with the last digit minus 2 (eg 252, 124, etc.).

Management address: IP address used to manage the device, in the in-band scenario。

We are a 100G switch with Nos, 100G module/network card factory in Shenzhen, China. We can provide you with one-stop service on products, transportation, customs clearance, and tariffs.