Ctrlk
VitePress 镜像站QQ 群 787969044视频教程Ⅰ视频教程Ⅱ
discord
For the complete documentation index, see llms.txt. This page is also available as Markdown.

7.1 Computer Networking Basics

FreeBSD is an ideal operating system for internet or intranet servers, capable of providing robust network services under extremely high loads, efficiently using memory, and maintaining good response times with thousands of concurrent user processes.

FreeBSD network configuration involves several core commands and configuration files:

Command/File
Purpose

ifconfig

Configure network interface parameters

route

Manually manipulate network routing tables

/etc/rc.conf

The core file for system startup configuration; persistent configurations for network interfaces are stored in this file

/etc/resolv.conf

Configure DNS resolver information

/etc/hosts

Provide static mapping of local hostnames to IP addresses

Network Model Basics

A computer network is a system that connects multiple computers and their peripheral devices, which are located in different geographical positions and have independent functions, through communication lines. Under the management and coordination of network operating systems, network management software, and network communication protocols, this network enables resource sharing and information transmission.

The FreeBSD network subsystem is implemented based on the TCP/IP protocol suite, following the layered architecture of the Internet protocol suite.

The TCP/IP protocol suite uses a four-layer model to organize network functions:

Layer
Name
Description

Layer 1

Link Layer

Responsible for sending and receiving data frames on physical network media, handling hardware address (MAC address) resolution. This layer corresponds to network interface drivers in FreeBSD, whose configuration is managed by the ifconfig command. Ethernet frame transmission and reception, and address resolution (ARP, Address Resolution Protocol) are all completed at this layer

Layer 2

Internet Layer

Responsible for routing and forwarding of data packets. The core protocol is IP (Internet Protocol), which also handles logical addressing (IP addresses), fragmentation and reassembly, and route selection. Routing table management and route command operations all belong to this layer

Layer 3

Transport Layer

Provides end-to-end communication services. The main protocols are TCP (Transmission Control Protocol) and UDP (User Datagram Protocol). TCP provides reliable connection-oriented transmission, while UDP provides connectionless unreliable transmission. FreeBSD implements a multi-TCP stack coexistence architecture, allowing the system to load multiple TCP protocol stack implementations simultaneously

Layer 4

Application Layer

Contains various user-oriented network application protocols, such as HTTP, SSH, DNS, SMTP, etc. FreeBSD provides a large number of network service software through Ports and pkg

References

  • Kurose J F, Ross K W. Computer Networking: A Top-Down Approach (8th Edition)[M]. Translated by Chen Ming. Beijing: China Machine Press, 2022. ISBN: 978-7-111-71236-7.

Identifying Network Adapters

FreeBSD supports a variety of wired and wireless network adapters. Check the hardware compatibility list for the FreeBSD version in use to confirm whether the network adapter is supported.

Identifying Network Adapters via the pciconf Command

To obtain the network adapters used by the system, execute the following command:

Example output is as follows:

The text before the @ symbol is the name of the driver that controls the device. In this example, they are em(4) and iwm(4), respectively.

Identifying Network Adapters via the ifconfig Command

Use the ifconfig command to view the list of network interfaces and their status on the system. The output should be similar to the following:

FreeBSD names network interfaces using the driver name followed by a unit number. The unit number indicates the order in which the adapter was detected at boot time, or the order in which it was subsequently detected. For example, em0 is the first network card in the system using the em(4) driver, and wlan0 is the first wireless interface created using the rtwn(4) driver.

This example shows the following devices:

  • em0: Ethernet interface.

  • lo0: Local loopback interface, used for internal communication within the local machine, not a physical network card. It can be used for performance analysis, software testing, and local communication.

  • wlan0: Generic WiFi 802.11 link layer interface, used for connecting to wireless networks.

    Tip

    If the ifconfig output only shows the lo0 interface, it means the system has not recognized any physical network card. In this case, check the network card hardware connection and driver loading status. You can use the dmesg | grep ether command to view the network card driver loading log.

This example shows that em0 and wlan0 are up and running normally.

Key indicator items:

Indicator
Description

UP

Indicates that the interface has been enabled by the administrator (in the up state). Currently, both em0 and wlan0 are enabled

inet (IPv4 address)

The address of the wired interface em0 is updated to 192.168.5.22; the address of the wireless interface wlan0 is 192.168.5.23

inet6 (IPv6 address)

em0 currently has two IPv6 addresses: the link-local address fe80::20c:29ff:fe84:f86%em0 and the global dynamic address 240e:341:207:a600:2d60:b653:3a68:8605

netmask (subnet mask)

The masks for both em0 and wlan0 are 0xffffff00, which is equivalent to the standard 255.255.255.0

broadcast (broadcast address)

The valid broadcast addresses for the network segments where both interfaces are currently located are 192.168.5.255

ether (MAC address)

The physical address of the wired interface em0 is 00:0c:29:84:0f:86; the physical address of the wireless interface wlan0 is 20:0d:b0:c4:ab:59

media (physical media)

em0 shows Ethernet autoselect mode (Ethernet autoselect (1000baseT <full-duplex>)); wlan0 shows wireless high-speed mode (IEEE 802.11 Wireless Ethernet VHT mode 11ac)

status (link status)

The status of em0 is active, indicating that the network cable is connected and the carrier signal is normal; the status of wlan0 is associated, indicating that it has successfully associated with the wireless access point (SSID externally displayed as test_5G)

In addition, the key indicator items for wlan0 are as follows:

Indicator
Description

ssid (wireless network name)

The name of the currently connected Wi-Fi is test_5G

channel (operating channel and frequency)

Currently operating on channel 153, with a frequency of 5765 MHz (in the 5GHz band), and 80MHz bandwidth enabled (vht/80-)

bssid (wireless router MAC address)

The physical address of the currently connected Wi-Fi hotspot is e4:60:4d:97:00:e8

country / regdomain (country region code and wireless regulatory domain)

Currently complying with US standards (country US) and the radio regulations of the US Federal Communications Commission (regdomain FCC)

authmode / privacy (authentication and encryption mode)

The authentication method is WPA2/802.11i (currently the mainstream secure wireless authentication standard), and privacy encryption is enabled (privacy ON). Unicast and transmission keys use AES-CCM (128-bit)

txpower (transmit power)

The current wireless transmit power is 17 dBm

If the ifconfig(8) output is similar to the following, it indicates that the network interface is pending configuration:

Gateways and Routing

Routing is the mechanism by which a system finds network paths to other systems. Each route is defined by a pair of addresses, representing the "destination" and the "gateway," respectively. A route indicates that when connecting to a specified destination, data packets should be sent through the specified gateway. Destinations are divided into three types: a single host, a subnet, and "default." The "default route" takes effect when no other route is applicable. Gateways are also divided into three types: a single host, an interface (also called a link), and an Ethernet hardware (MAC) address. Known routes are stored in the routing table.

Routing Basics

Use netstat(1) to view the routing table of a FreeBSD system. Adding the -n option avoids reverse DNS resolution delays, which is particularly important when troubleshooting network issues:

The entries in this example are as follows:

  • default: The default route. When the local system needs to connect to a remote host, it checks the routing table to determine whether a known path exists. If the remote host matches an entry in the table, the system checks whether the connection can be completed through the interface specified by that entry. The default route is used when there is no more specific prefix match, and is not affected by the status of other paths. For a host in a local area network, the Gateway field of the default route should be set to a system that can directly connect to the internet. When reading this entry, confirm that the Flags column indicates the gateway is reachable (UG).

    For a host acting as an external gateway, its default route points to the gateway connected to the Internet Service Provider (ISP).

    In the IPv4 network, the default gateway is 192.168.179.2, marked as UGS (Up, Gateway, Static). All IPv4 traffic destined for unknown external networks (such as the internet) is sent through the em0 interface to this router for forwarding.

    In the IPv6 section, the default route specifies the next hop as the link-local address fe80::5%em1, which is typically a router provided by the ISP. The flag is UG, and all unmatched IPv6 traffic is sent out through the em1 interface. Link-local addresses are an IPv6 feature, used only for local link communication, and are not equivalent to publicly reachable addresses.

    This machine has dual-stack network capability, with IPv4 and IPv6 egress depending on different network cards (em0 and em1), respectively.

  • localhost: The second route. The interface specified for localhost in the Netif column is lo0, also known as the loopback device. All traffic sent to this destination remains within the local machine and is not sent over the network.

  • Subnets

    The IPv4 routing table shows that this device is connected to two local area networks simultaneously: 192.168.5.0/24 and 192.168.179.0/24. The corresponding gateways are link#2 and link#1, respectively, with flags containing only U (Up), indicating that both are "directly connected networks." When this machine communicates with other devices within these two network segments (such as computers in the 192.168.5.x range), data packets do not need to pass through a router; instead, after resolving the target MAC address via ARP, they are sent directly as link-layer unicast frames through the em1 and em0 interfaces.

  • Global Unicast Addresses

    In the IPv6 routing table (Internet6), entries starting with 240e:341:22b:ae00: are bound to the em1 interface. This address belongs to a Global Unicast Address, typically assigned by an ISP (such as China Telecom, starting with 240e), and is theoretically routable on the internet, but actual accessibility depends on firewall policies and service configurations.

  • Flags

    The Flags column displays the attributes of each route. The following table summarizes common routing table flags and their meanings.

    Common Routing Table Flags

    Flag
    Purpose

    U

    Route is up

    H

    Route target is a single host

    G

    Forward any traffic for this destination to this gateway, and the gateway decides how to further forward it

    S

    This route is statically configured

    M

    Route has been modified by a redirect

    B

    Blackhole route: silently discard matching packets

    D

    Dynamically created by a redirect

    L

    Link-layer route, involving Ethernet hardware address reference

    R

    Reject route: destination host or network is unreachable

    b

    Route represents a broadcast address

Tracing Route Information

After address space is allocated to a network, the service provider configures its own routing table to ensure that all traffic is sent to the link for that site. How do external sites determine that data packets should be sent to the ISP for that network?

The internet tracks all allocated address space through a global routing system and defines its connection nodes to the internet backbone (the core lines that carry internet traffic). Each backbone router maintains a master routing table that directs traffic for specific networks to the corresponding backbone carrier, and then passes it through a series of service providers tier by tier to the destination network.

Service providers must announce their connection points to backbone sites so that traffic can reach that network. This process is called route propagation.

Route propagation occasionally experiences failures, causing some sites to become unreachable. In such cases, the commonly used command to find the routing break point is traceroute (use traceroute6 for IPv6), which is particularly useful when ping fails.

Using traceroute requires providing the address of the remote host. The output will display the gateway hosts along the path, ultimately reaching the target host or terminating due to a connection interruption.

The following is an IPv6 route trace to the freebsd.org domain:

Detailed explanation:

No.
Node
Description

1

*

No response from the first hop of the local network

2

240e:c::201 (240e:c::201)

ISP access gateway

3, 4

*

No response from backbone pre-level nodes, possibly dropping ICMPv6

5

240e::1:11:46:5c02 (240e::1:11:46:5c02)

China Telecom core backbone router, domestic long-haul aggregation node

6

240e::f:1:6601:503 (240e::f:1:6601:503)

China Telecom backbone node before the international exit, some probe packets lost

7

240e:0:a::c9:360d (240e:0:a::c9:360d), 240e:0:a::c9:3649 (240e:0:a::c9:3649)

China Telecom intercontinental backbone routers for overseas exit

8

*

No response node, possibly dropping ICMPv6

9

zayo.ae10.mpr4.sjc7.us.zip.zayo.com (2001:438:ffff::407e:2f9)

Zayo San Jose, California, USA backbone network (Zip Zayo) entry node

10

ae34.mpr1.ewr4.us.zip.zayo.com (2001:438:ffff::407d:1455)

Zayo Newark, USA backbone network node, responsible for East Coast traffic aggregation

11

2001:438:fffe::24ba (2001:438:fffe::24ba)

Zayo US backbone last hop, approaching NYInternet entry

12

cs89-cs80.nyinternet.net (2610:1c1::2502)

NYInternet US carrier core router, connecting to the freebsd.org network

13

2610:1c1::803 (2610:1c1::803)

NYInternet core backbone node, responsible for traffic distribution to freebsd.org servers

14

wfe0.nyi.freebsd.org (2610:1c1:1:606c::50:15)

freebsd.org server terminal node, target of traceroute6

Previous6.6 Compression and ArchivingNext7.2 Basic Network Management

Last updated