Router selection protocol for the advanced mode of the Network Gateway

Router selection protocol for the advanced mode of the Network Gateway

The most fundamental starting point of routing for routing services:
We need the hierarchical logical management of networks. Therefore, we need an IPv4 Network to allow communication between networks )".
Design Network:
1) define requirements;
2) Deploy routes;
Branch step arrangement:
1) create a deployment plan;
2) Deploy the solution;
3) Documentation
Deployment steps:
1) preparation stage
2) planning stage
3) design phase
4) implementation phase
5) operation phase
6) optimization stage
(Enhanced Internal Gateway Routing Protocol)
The following are the advantages of network replication:
L advanced distance vector
L update is sent using multicast and unicast instead of broadcast.
L support multiple network layer protocols
L 100% non-ring, non-category Routing
L fast convergence
L trigger incremental update
L flexible network design
L "Server Load balancer" that supports other overhead and non-other overhead"
L The configuration for multiple network environments is still "simple"
L support manual summarization at any point
L use composite metric values
The key features of the kernel: (the kernel of the kernel is encapsulated by IP addresses)
L Neighbor Discovery and recovery mechanism
L reliable transmission protocol
L Protocol Independent Unit
L DUAL Algorithm
In the Protocol field of the IP packet, "88 and 89" is marked in the protocol field of the IP packet ".
The Hello packet is used to establish the neighbor. The Hello packet is sent through the address 224.0.0.10.
 
There are five packet types in the network:
1) Hello (Establish and maintain a neighbor relationship)
2) Updata (update package, send route update)
3) Query (Query package, used to send Query information about a route to a neighbor .)
4) Reple (response packet, which responds to queries .)
5) ACK (confirm the package and respond to the reliable package .)
 
Initial route discovery:

First, R1 sends the Hello packet on the interface that enables the replication gateway. After receiving the Hello packet, R2 also sends the Hello packet to the other party and sends its complete route information to the other party through Updata, because Updata is reliable, R1 will confirm the ACK after receiving it, and then put the received route information in its own Topolegy table. Then, it will send its route information to R2 through the Updata package, r2 receives the route information sent by R1, returns the ACK message, and puts the route information in its own Topology.
 
The following table lists the three tables of the source node:
1) neighbor table
The neighbor table only stores the directly connected vrouters that run this VPN gateway.
2) topology table
How far is AD (advertising distance) from my neighbor to the target.
FD (feasible distance) is how far I am from my neighbor plus my neighbor to the target.
Successor (Successor router), the smallest FD becomes Successor and will be placed in the routing table.
Feasible Successor (Feasible Successor router): Its AD must be less than the FD of Successor to become Feasible Successor. Place the Feasible Successor in the "topology table.
3) route table
Successor and Feasible Successor ":
To become a Feasible Successor, His "AD distance" must be a "Feasible distance (FD)" from Successor, which effectively avoids the generation of "loops!
 
Metric (Metric value) of MongoDB ):
Bandwidth (Bandwidth) -- (K1) Bandwidth is the minimum Bandwidth of the inbound interface updated along the way. Divide the Bandwidth by 10 ^ 7.
Delay (latency) -- (K3) latency is the sum of the latency of update inbound interfaces along the way.
Reliability -- (K4 AND K5)
Loading (load) -- (K2)
MTU is not included in the calculation, but is included!
Default metric value = [bandwidth + latency] X 256 (The unit of latency is 10 usec, so the latency is divided by 10 .)