For example:

• Subnet mask notation: 255.255.255.0
• CIDR notation: 192.168.1.0/24

Both expressions serve the purpose of distinguishing between the network and host parts, but CIDR provides a more flexible method of network scoping, which can reduce IP address waste.

IPv4 CIDR

IPv6 CIDR

In IPv6, the method of allocating subnets and network size is very different from IPv4. Typically, the /64 subnet is used as a standard, which allows 2\^64 host addresses. Below is a table showing the major subnet notations and number of hosts in IPv6. In IPv6, CIDR notation is used rather than the concept of a subnet mask, and there is no direct correspondence between netmask and Cisco mask like in IPv4.

Key points

• /64: Most IPv6 subnets are assigned /64. It can provide addresses to countless individual devices on a single network.
• /127: Used for point-to-point links in IPv6, only two hosts are connected.
• /128: Used to assign a specific IP address to a single device or host.

In IPv6, instead of using various subnet sizes to differentiate networks like IPv4, most subnets are set to /64 to maintain consistency in address allocation.

International standard special band

L3 allocation rules

1. “Absolutely reserved addresses” that cannot be used as hosts (2)

The start and end addresses of any subnet cannot be assigned to devices (PCs, servers, etc.).

• Network Address: First address of the band (e.g. 192.168.0.0)
• Purpose: A name tag that represents the network itself. When the router finds a route, it looks at this address and decides, “Oh, I can send this packet to the 192.168.0 point neighborhood.”
• Broadcast Address: Last address of the band (e.g. 192.168.0.255)
• Purpose: Used to send data to all devices at once belonging to the network. If you send a packet to this address, all PCs in the network will receive its contents.

2. Customarily used address (Gateway)

Technically, assignment is possible, but for management efficiency, these addresses are left blank as a promise.

• Default Gateway: Usually .1 or .254
• Purpose: This is the address of the router or router that is the ‘entrance’ to the external Internet.
• Features: Although not established by law, it is a strong commitment among network engineers. Usually the first address (.1) is used the most, followed by the last available address (.254).
• Server and Fixed Equipment Band: Mainly .2 to .10
• For management convenience, the numbers right behind the gateway are often statically assigned to major infrastructure equipment such as DNS servers, firewalls, and NAS.

3. Exception rules: /32, /31

• /32: In case there is only one address. It is mainly used when setting the Loopback IP of the device. At this time, the network/broadcast distinction is meaningless, so the device uses the address itself.
• /31: Used to reduce IP waste in connections between modern router devices. (RFC 3021) Only in this case, as an exception, both IPs are assigned to the device without a network/broadcast address. However, it is never used in a typical PC or switch environment.