linux 网卡配置 vlan/bond/bridge/macvlan/ipvlan/macvtap 模式

天空闲话  金牌会员 | 2024-9-9 07:04:56 | 显示全部楼层 | 阅读模式
打印 上一主题 下一主题
楼主

主题 537|帖子 537|积分 1611

linux 网卡模式

linux网卡支持非vlan模式、vlan模式、bond模式、bridge模式，macvlan模式、ipvlan模式等，下面介绍互换机端及服务器端配置示例。
前置要求：


  • 预备一台物理互换机，以 H3C S5130 三层互换机为例
  • 预备一台物理服务器，以 Ubuntu 22.04 LTS 操作体系为例
互换机创建2个示例VLAN，vlan10和vlan20，及VLAN接口。
  1. <H3C>system-view
  3. [H3C]vlan 10 20
  5. [H3C]interface Vlan-interface 10
  6. [H3C-Vlan-interface10]ip address 172.16.10.1 24
  7. [H3C-Vlan-interface10]undo shutdown
  8. [H3C-Vlan-interface10]exit
  9. [H3C]
  11. [H3C]interface Vlan-interface 20
  12. [H3C-Vlan-interface20]ip address 172.16.20.1 24
  13. [H3C-Vlan-interface20]undo  shutdown
  14. [H3C-Vlan-interface20]exit
  15. [H3C]
复制代码
网卡非vlan模式

网卡非vlan模式，一样平常直接配置IP地点，对端上连互换机配置为access口，access口一样平常用于连接纯物理服务器或办公终端设备。
示意图如下

互换机配置，互换机接口配置为access模式，并参加对应vlan
  1. <H3C>system-view
  2. [H3C]interface GigabitEthernet 1/0/1
  3. [H3C-GigabitEthernet1/0/1]port link-type access
  4. [H3C-GigabitEthernet1/0/1]port access vlan 10
  5. [H3C-GigabitEthernet1/0/1]exit
  6. [H3C]
  7. [H3C]interface GigabitEthernet 1/0/2
  8. [H3C-GigabitEthernet1/0/2]port link-type access
  9. [H3C-GigabitEthernet1/0/2]port access vlan 20
  10. [H3C-GigabitEthernet1/0/2]exit
  11. [H3C]
复制代码
服务器1配置，服务器网卡直接配置IP地点
  1. root@server1:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   ethernets:
  4.     enp1s0:
  5.       dhcp4: false
  6.       addresses:
  7.         - 172.16.10.10/24
  8.       nameservers:
  9.         addresses:
  10.           - 223.5.5.5
  11.           - 223.6.6.6
  12.       routes:
  13.         - to: default
  14.           via: 172.16.10.1
  15.   version: 2
复制代码
服务器2配置，服务器网卡直接配置IP地点
  1. root@server2:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   ethernets:
  4.     enp1s0:
  5.       dhcp4: false
  6.       addresses:
  7.         - 172.16.20.10/24
  8.       nameservers:
  9.         addresses:
  10.           - 223.5.5.5
  11.           - 223.6.6.6
  12.       routes:
  13.         - to: default
  14.           via: 172.16.20.1
  15.   version: 2
复制代码
应用网卡配置
  1. netplan apply
复制代码
查看服务器接口信息
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
  9.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  10.     inet 172.16.10.10/24 brd 172.16.10.255 scope global enp1s0
  11.        valid_lft forever preferred_lft forever
  12.     inet6 fe80::7eb5:9bff:fe59:a71/64 scope link
  13.        valid_lft forever preferred_lft forever
复制代码
通过server1 ping server2测试连通性，三层互换机支持路由功能，可以或许打通二层隔离的vlan网段。
  1. root@server1:~# ping 172.16.20.10 -c 4
  2. PING 172.16.20.10 (172.16.20.10) 56(84) bytes of data.
  3. 64 bytes from 172.16.20.10: icmp_seq=1 ttl=64 time=0.033 ms
  4. 64 bytes from 172.16.20.10: icmp_seq=2 ttl=64 time=0.048 ms
  5. 64 bytes from 172.16.20.10: icmp_seq=3 ttl=64 time=0.048 ms
  6. 64 bytes from 172.16.20.10: icmp_seq=4 ttl=64 time=0.047 ms
  8. --- 172.16.20.10 ping statistics ---
  9. 4 packets transmitted, 4 received, 0% packet loss, time 3061ms
  10. rtt min/avg/max/mdev = 0.033/0.044/0.048/0.006 ms
复制代码
网卡vlan模式

vlan模式下，对端上连互换机须要配置为trunk口，允许多个vlan通过。
示意图如下

互换机配置，互换机须要配置为trunk口，允许多个vlan通过
  1. H3C>system-view
  2. [H3C]interface GigabitEthernet 1/0/1
  3. [H3C-GigabitEthernet1/0/1]port link-type trunk
  4. [H3C-GigabitEthernet1/0/1]port trunk permit vlan 10 20
  5. [H3C-GigabitEthernet1/0/1]exit
  6. [H3C]
复制代码
服务器配置，服务器须要配置vlan子接口
  1. root@server1:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   ethernets:
  4.     enp1s0:
  5.       dhcp4: true
  6.   vlans:
  7.     vlan10:
  8.       id: 10
  9.       link: enp1s0
  10.       addresses: [ "172.16.10.10/24" ]
  11.       routes:
  12.         - to: default
  13.           via: 172.16.10.1
  14.           metric: 200
  15.     vlan20:
  16.       id: 20
  17.       link: enp1s0
  18.       addresses: [ "172.16.20.10/24" ]
  19.       routes:
  20.         - to: default
  21.           via: 172.16.20.1
  22.           metric: 300
  23.   version: 2
复制代码
查看接口信息，新建了两个vlan子接口vlan10和vlan20
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
  9.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  10.     inet6 fe80::7eb5:9bff:fe59:a71/64 scope link
  11.        valid_lft forever preferred_lft forever
  12. 10: vlan10@enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  13.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  14.     inet 172.16.10.10/24 brd 172.16.10.255 scope global vlan10
  15.        valid_lft forever preferred_lft forever
  16.     inet6 fe80::7eb5:9bff:fe59:a71/64 scope link
  17.        valid_lft forever preferred_lft forever
  18. 11: vlan20@enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  19.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  20.     inet 172.16.20.10/24 brd 172.16.20.255 scope global vlan20
  21.        valid_lft forever preferred_lft forever
  22.     inet6 fe80::7eb5:9bff:fe59:a71/64 scope link
  23.        valid_lft forever preferred_lft forever
复制代码
通过vlan10 和 vlan20测试与网关连通性
  1. root@server1:~# ping 172.16.10.1 -c 4
  2. PING 172.16.10.1 (172.16.10.1) 56(84) bytes of data.
  3. 64 bytes from 172.16.10.1: icmp_seq=1 ttl=64 time=0.033 ms
  4. 64 bytes from 172.16.10.1: icmp_seq=2 ttl=64 time=0.048 ms
  5. 64 bytes from 172.16.10.1: icmp_seq=3 ttl=64 time=0.048 ms
  6. 64 bytes from 172.16.10.1: icmp_seq=4 ttl=64 time=0.047 ms
  8. --- 172.16.10.1 ping statistics ---
  9. 4 packets transmitted, 4 received, 0% packet loss, time 3061ms
  10. rtt min/avg/max/mdev = 0.033/0.044/0.048/0.006 ms
  11. root@server1:~#
  12. root@server1:~# ping 172.16.20.1 -c 4
  13. PING 172.16.20.1 (172.16.20.1) 56(84) bytes of data.
  14. 64 bytes from 172.16.20.1: icmp_seq=1 ttl=64 time=0.033 ms
  15. 64 bytes from 172.16.20.1: icmp_seq=2 ttl=64 time=0.048 ms
  16. 64 bytes from 172.16.20.1: icmp_seq=3 ttl=64 time=0.048 ms
  17. 64 bytes from 172.16.20.1: icmp_seq=4 ttl=64 time=0.047 ms
  19. --- 172.16.20.1 ping statistics ---
  20. 4 packets transmitted, 4 received, 0% packet loss, time 3061ms
  21. rtt min/avg/max/mdev = 0.033/0.044/0.048/0.006 ms
复制代码
网卡bond模式

bond模式下，对端互换机须要配置bond聚合口。
示意图如下

互换机配置，配置动态链路聚合，将端口g1/0/1和g1/0/3参加聚合组。然后将bond口配置为trunk模式。
  1. <H3C>system-view
  2. [H3C]interface Bridge-Aggregation 1
  3. [H3C-Bridge-Aggregation1]link-aggregation mode dynamic
  4. [H3C-Bridge-Aggregation1]quit
  6. [H3C]interface GigabitEthernet 1/0/1
  7. [H3C-GigabitEthernet1/0/1]port link-aggregation group 1
  8. [H3C-GigabitEthernet1/0/1]exit
  10. [H3C]interface GigabitEthernet 1/0/3
  11. [H3C-GigabitEthernet1/0/3]port link-aggregation group 1
  12. [H3C-GigabitEthernet1/0/3]exit
  14. [H3C]interface Bridge-Aggregation 1
  15. [H3C-Bridge-Aggregation1]port link-type trunk
  16. [H3C-Bridge-Aggregation1]port trunk permit vlan 10 20
  17. [H3C-Bridge-Aggregation1]exit
复制代码
服务器配置
  1. root@server1:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   version: 2
  4.   ethernets:
  5.     enp1s0:
  6.       dhcp4: no
  7.     enp2s0:
  8.       dhcp4: no
  9.   bonds:
  10.     bond0:
  11.       interfaces:
  12.         - enp1s0
  13.         - enp2s0
  14.       parameters:
  15.         mode: 802.3ad
  16.         lacp-rate: fast
  17.         mii-monitor-interval: 100
  18.         transmit-hash-policy: layer2+3
  19.   vlans:
  20.     vlan10:
  21.       id: 10
  22.       link: bond0
  23.       addresses: [ "172.16.10.10/24" ]
  24.       routes:
  25.         - to: default
  26.           via: 172.16.10.1
  27.           metric: 200
  28.     vlan20:
  29.       id: 20
  30.       link: bond0
  31.       addresses: [ "172.16.20.10/24" ]
  32.       routes:
  33.         - to: default
  34.           via: 172.16.20.1
  35.           metric: 300
复制代码
查看网卡信息，新建了bond0网口，而且基于bond0网口创建了两个vlan子接口vlan10和vlan20，enp1s0和enp2s0体现master bond0，阐明两个网卡属于bond0成员接口。
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc fq_codel master bond0 state UP group default qlen 1000
  9.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff permaddr 7c:b5:9b:59:0a:71
  10. 3: enp2s0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc fq_codel master bond0 state UP group default qlen 1000
  11.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff permaddr e4:54:e8:dc:e5:88
  12. 7: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  13.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff
  14.     inet6 fe80::acfd:60ff:fe48:841a/64 scope link
  15.        valid_lft forever preferred_lft forever
  16. 8: vlan10@bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  17.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff
  18.     inet 172.16.10.10/24 brd 172.16.10.255 scope global vlan10
  19.        valid_lft forever preferred_lft forever
  20.     inet6 fe80::acfd:60ff:fe48:841a/64 scope link
  21.        valid_lft forever preferred_lft forever
  22. 9: vlan20@bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  23.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff
  24.     inet 172.16.20.10/24 brd 172.16.20.255 scope global vlan20
  25.        valid_lft forever preferred_lft forever
  26.     inet6 fe80::acfd:60ff:fe48:841a/64 scope link
  27.        valid_lft forever preferred_lft forever
复制代码
查看bond状态，Bonding Mode体现为IEEE 802.3ad Dynamic link aggregation，而且下面Slave Interface体现了两个成员接口的信息。
  1. root@server1:~# cat /proc/net/bonding/bond0
  2. Ethernet Channel Bonding Driver: v5.15.0-60-generic
  4. Bonding Mode: IEEE 802.3ad Dynamic link aggregation
  5. Transmit Hash Policy: layer2+3 (2)
  6. MII Status: up
  7. MII Polling Interval (ms): 100
  8. Up Delay (ms): 0
  9. Down Delay (ms): 0
  10. Peer Notification Delay (ms): 0
  12. 802.3ad info
  13. LACP active: on
  14. LACP rate: fast
  15. Min links: 0
  16. Aggregator selection policy (ad_select): stable
  17. System priority: 65535
  18. System MAC address: ae:fd:60:48:84:1a
  19. Active Aggregator Info:
  20.         Aggregator ID: 1
  21.         Number of ports: 2
  22.         Actor Key: 9
  23.         Partner Key: 1
  24.         Partner Mac Address: fc:60:9b:35:ad:18
  26. Slave Interface: enp1s0
  27. MII Status: up
  28. Speed: 1000 Mbps
  29. Duplex: full
  30. Link Failure Count: 2
  31. Permanent HW addr: 7c:b5:9b:59:0a:71
  32. Slave queue ID: 0
  33. Aggregator ID: 1
  34. Actor Churn State: none
  35. Partner Churn State: none
  36. Actor Churned Count: 0
  37. Partner Churned Count: 0
  38. details actor lacp pdu:
  39.     system priority: 65535
  40.     system mac address: ae:fd:60:48:84:1a
  41.     port key: 9
  42.     port priority: 255
  43.     port number: 1
  44.     port state: 63
  45. details partner lacp pdu:
  46.     system priority: 32768
  47.     system mac address: fc:60:9b:35:ad:18
  48.     oper key: 1
  49.     port priority: 32768
  50.     port number: 2
  51.     port state: 61
  53. Slave Interface: enp2s0
  54. MII Status: up
  55. Speed: 1000 Mbps
  56. Duplex: full
  57. Link Failure Count: 3
  58. Permanent HW addr: e4:54:e8:dc:e5:88
  59. Slave queue ID: 0
  60. Aggregator ID: 1
  61. Actor Churn State: none
  62. Partner Churn State: none
  63. Actor Churned Count: 0
  64. Partner Churned Count: 0
  65. details actor lacp pdu:
  66.     system priority: 65535
  67.     system mac address: ae:fd:60:48:84:1a
  68.     port key: 9
  69.     port priority: 255
  70.     port number: 2
  71.     port state: 63
  72. details partner lacp pdu:
  73.     system priority: 32768
  74.     system mac address: fc:60:9b:35:ad:18
  75.     oper key: 1
  76.     port priority: 32768
  77.     port number: 1
  78.     port state: 61
复制代码
测试连通性，测试与互换机网关地点的连通性：
  1. root@server1:~# ping 172.16.10.1 -c 4
  2. PING 172.16.10.1 (172.16.10.1) 56(84) bytes of data.
  3. 64 bytes from 172.16.10.1: icmp_seq=1 ttl=255 time=1.64 ms
  4. 64 bytes from 172.16.10.1: icmp_seq=2 ttl=255 time=1.59 ms
  5. 64 bytes from 172.16.10.1: icmp_seq=3 ttl=255 time=1.95 ms
  6. 64 bytes from 172.16.10.1: icmp_seq=4 ttl=255 time=1.93 ms
  8. --- 172.16.10.1 ping statistics ---
  9. 4 packets transmitted, 4 received, 0% packet loss, time 3006ms
  10. rtt min/avg/max/mdev = 1.589/1.776/1.953/0.165 ms
  11. root@server1:~#
复制代码
关闭一个接口，再次测试连通性，依然可以或许ping通
  1. root@server1:~# ip link set dev enp2s0 down
  2. root@server1:~# ip link show enp2s0
  3. 3: enp2s0: <BROADCAST,MULTICAST,SLAVE> mtu 1500 qdisc fq_codel master bond0 state DOWN mode DEFAULT group default qlen 1000
  4.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff permaddr e4:54:e8:dc:e5:88
  5. root@server1:~#
  6. root@server1:~# ping 172.16.10.1 -c 4
  7. PING 172.16.10.1 (172.16.10.1) 56(84) bytes of data.
  8. 64 bytes from 172.16.10.1: icmp_seq=1 ttl=255 time=1.54 ms
  9. 64 bytes from 172.16.10.1: icmp_seq=2 ttl=255 time=1.64 ms
  10. 64 bytes from 172.16.10.1: icmp_seq=3 ttl=255 time=2.73 ms
  11. 64 bytes from 172.16.10.1: icmp_seq=4 ttl=255 time=1.47 ms
  13. --- 172.16.10.1 ping statistics ---
  14. 4 packets transmitted, 4 received, 0% packet loss, time 3006ms
  15. rtt min/avg/max/mdev = 1.470/1.844/2.732/0.516 ms
复制代码
网卡桥接模式

桥接模式下，对端互换机可配置access模式或trunk模式。
示意图如下

互换机配置，互换机接口配置为access模式为例，并参加对应vlan
  1. <H3C>system-view
  2. [H3C]interface GigabitEthernet 1/0/1
  3. [H3C-GigabitEthernet1/0/1]port link-type access
  4. [H3C-GigabitEthernet1/0/1]port access vlan 10
  5. [H3C-GigabitEthernet1/0/1]exit
  6. [H3C]
复制代码
服务器配置，物理网卡参加到网桥中，IP地点配置到网桥接口br0上。
  1. root@server1:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   version: 2
  4.   ethernets:
  5.     enp1s0:
  6.       dhcp4: no
  7.       dhcp6: no
  8.   bridges:
  9.     br0:
  10.       interfaces: [enp1s0]
  11.       addresses: [172.16.10.10/24]
  12.       routes:
  13.       - to: default
  14.         via: 172.16.10.1
  15.         metric: 100
  16.         on-link: true
  17.       mtu: 1500
  18.       nameservers:
  19.         addresses:
  20.           - 223.5.5.5
  21.           - 223.6.6.6
  22.       parameters:
  23.         stp: true
  24.         forward-delay: 4
复制代码
查看网卡信息
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel master br0 state UP group default qlen 1000
  9.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  10. 12: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  11.     link/ether 0e:d0:7e:31:9c:74 brd ff:ff:ff:ff:ff:ff
  12.     inet 172.16.10.10/24 brd 172.16.10.255 scope global br0
  13.        valid_lft forever preferred_lft forever
  14.     inet6 fe80::cd0:7eff:fe31:9c74/64 scope link
  15.        valid_lft forever preferred_lft forever
复制代码
查看网桥及接口，当前网桥上只有一个物理接口enp1s0。
  1. root@server1:~# apt install -y bridge-utils
  2. root@ubuntu:~# brctl show
  3. bridge name     bridge id               STP enabled     interfaces
  4. br0             8000.0ed07e319c74       yes             enp1s0
  5. root@server1:~#
复制代码
这样在KVM捏造化情况，捏造机实例连接到网桥后，捏造机可以配置与物理网卡类似网段的IP地点。访问捏造机可以像访问物理机一样方便。
网卡macvlan模式

macvlan(MAC Virtual LAN)是Linux内核提供的一种网络捏造化技术，它允许在一个物理网卡接口上创建多个捏造网卡接口，每个捏造接口都有自己独立的MAC地点，也可以配置上 IP 地点进行通信。Macvlan 下的捏造机或者容器网络和主机在同一个网段中，共享同一个广播域。
macvlan模式下，对端互换机可配置access模式或trunk模式，trunk模式下macvlan可以或许与vlan很好的结合使用。
示意图如下：

macvlan IP模式

该模式下，上连互换机接口配置为access模式，服务器macvlan主网卡和子接口直接配置类似网段的IP地点。
互换机配置
  1. <H3C>system-view
  2. [H3C]interface GigabitEthernet 1/0/1
  3. [H3C-GigabitEthernet1/0/1]port link-type access
  4. [H3C-GigabitEthernet1/0/1]port access vlan 10
  5. [H3C-GigabitEthernet1/0/1]exit
复制代码
服务器配置，macvlan支持多种模式，这里使用bridge模式，并长期化配置
  1. cat >/etc/networkd-dispatcher/routable.d/10-macvlan-interfaces.sh<<EOF
  2. #! /bin/bash
  3. ip link add macvlan0 link enp1s0 type macvlan mode bridge
  4. ip link add macvlan1 link enp1s0 type macvlan mode bridge
  5. EOF
  7. chmod o+x,g+x,u+x /etc/networkd-dispatcher/routable.d/10-macvlan-interfaces.sh
复制代码
配置netplan
  1. root@server1:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   ethernets:
  4.     enp1s0:
  5.       dhcp4: false
  6.       addresses:
  7.         - 172.16.10.10/24
  8.       nameservers:
  9.         addresses:
  10.           - 223.5.5.5
  11.           - 223.6.6.6
  12.       routes:
  13.         - to: default
  14.           via: 172.16.10.1
  15.     macvlan0:
  16.       addresses:
  17.         - 172.16.10.11/24
  18.     macvlan1:
  19.       addresses:
  20.         - 172.16.10.12/24
  21.   version: 2
复制代码
应用网卡配置
  1. netplan apply
复制代码
查看网卡信息，新建了两个macvlan接口，IP地点与主网卡位于同一网段，而且每个接口都有独立的MAC地点。
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
  9.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  10.     inet 172.16.10.10/24 brd 172.16.10.255 scope global enp1s0
  11.        valid_lft forever preferred_lft forever
  12. 13: macvlan0@enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  13.     link/ether 32:e8:b4:0a:47:62 brd ff:ff:ff:ff:ff:ff
  14.     inet 172.16.10.11/24 brd 172.16.10.255 scope global macvlan0
  15.        valid_lft forever preferred_lft forever
  16.     inet6 fe80::30e8:b4ff:fe0a:4762/64 scope link
  17.        valid_lft forever preferred_lft forever
  18. 14: macvlan1@enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  19.     link/ether d2:73:75:14:b2:04 brd ff:ff:ff:ff:ff:ff
  20.     inet 172.16.10.12/24 brd 172.16.10.255 scope global macvlan1
  21.        valid_lft forever preferred_lft forever
  22.     inet6 fe80::d073:75ff:fe14:b204/64 scope link
  23.        valid_lft forever preferred_lft forever
复制代码
测试与网关的连通性
  1. root@server1:~# ping -c 3 172.16.10.1
  2. PING 172.16.10.1 (172.16.10.1) 56(84) bytes of data.
  3. 64 bytes from 172.16.10.1: icmp_seq=1 ttl=255 time=3.60 ms
  4. 64 bytes from 172.16.10.1: icmp_seq=2 ttl=255 time=1.45 ms
  5. 64 bytes from 172.16.10.1: icmp_seq=3 ttl=255 time=1.44 ms
  7. --- 172.16.10.1 ping statistics ---
  8. 3 packets transmitted, 3 received, 0% packet loss, time 2004ms
  9. rtt min/avg/max/mdev = 1.441/2.163/3.602/1.017 ms
  10. root@server1:~#
复制代码
macvlan vlan模式

该模式下，上连互换机接口配置为trunk模式，服务器macvlan主网卡不配置IP地点，每个macvlan子接口配置为差别的vlan子接口。
互换机配置
  1. <H3C>system-view
  2. [H3C]interface GigabitEthernet 1/0/1
  3. [H3C-GigabitEthernet1/0/3]port link-type trunk
  4. [H3C-GigabitEthernet1/0/3]port trunk permit vlan 10 20
  5. [H3C-GigabitEthernet1/0/1]exit
  6. [H3C]
复制代码
服务器配置，macvlan支持多种模式，这里使用bridge模式，并长期化配置
  1. cat >/etc/networkd-dispatcher/routable.d/10-macvlan-interfaces.sh<<EOF
  2. #! /bin/bash
  3. ip link add macvlan0 link enp1s0 type macvlan mode bridge
  4. ip link add macvlan1 link enp1s0 type macvlan mode bridge
  5. EOF
  7. chmod o+x,g+x,u+x /etc/networkd-dispatcher/routable.d/10-macvlan-interfaces.sh
复制代码
配置netplan，两个macvlan接口macvlan0和macvlan1分别配置vlan子接口vlan10和vlan20。
  1. root@ubuntu:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   ethernets:
  4.     enp1s0:
  5.       dhcp4: false
  6.     macvlan0:
  7.       dhcp4: false
  8.     macvlan1:
  9.       dhcp4: false
  10.   vlans:
  11.     vlan10:
  12.       id: 10
  13.       link: macvlan0
  14.       addresses: [ "172.16.10.10/24" ]
  15.       routes:
  16.         - to: default
  17.           via: 172.16.10.1
  18.           metric: 200
  19.     vlan20:
  20.       id: 20
  21.       link: macvlan1
  22.       addresses: [ "172.16.20.10/24" ]
  23.       routes:
  24.         - to: default
  25.           via: 172.16.20.1
  26.           metric: 300
  27.   version: 2
复制代码
应用网卡配置
  1. netplan apply
复制代码
查看网卡信息，新建了两个macvlan接口，以及对应的两个vlan子接口。
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
  9.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  10.     inet6 fe80::7eb5:9bff:fe59:a71/64 scope link
  11.        valid_lft forever preferred_lft forever
  12. 11: macvlan0@enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  13.     link/ether 32:e8:b4:0a:47:62 brd ff:ff:ff:ff:ff:ff
  14.     inet6 fe80::30e8:b4ff:fe0a:4762/64 scope link
  15.        valid_lft forever preferred_lft forever
  16. 12: macvlan1@enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  17.     link/ether d2:73:75:14:b2:04 brd ff:ff:ff:ff:ff:ff
  18.     inet6 fe80::d073:75ff:fe14:b204/64 scope link
  19.        valid_lft forever preferred_lft forever
  20. 13: vlan10@macvlan0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  21.     link/ether 32:e8:b4:0a:47:62 brd ff:ff:ff:ff:ff:ff
  22.     inet 172.16.10.10/24 brd 172.16.10.255 scope global vlan10
  23.        valid_lft forever preferred_lft forever
  24.     inet6 fe80::30e8:b4ff:fe0a:4762/64 scope link
  25.        valid_lft forever preferred_lft forever
  26. 14: vlan20@macvlan1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  27.     link/ether d2:73:75:14:b2:04 brd ff:ff:ff:ff:ff:ff
  28.     inet 172.16.20.10/24 brd 172.16.20.255 scope global vlan20
  29.        valid_lft forever preferred_lft forever
  30.     inet6 fe80::d073:75ff:fe14:b204/64 scope link
  31.        valid_lft forever preferred_lft forever
复制代码
测试两个VLAN接口与外部网关的连通性
  1. root@server1:~# ping -c 3 172.16.10.1
  2. PING 172.16.10.1 (172.16.10.1) 56(84) bytes of data.
  3. 64 bytes from 172.16.10.1: icmp_seq=1 ttl=255 time=3.60 ms
  4. 64 bytes from 172.16.10.1: icmp_seq=2 ttl=255 time=1.45 ms
  5. 64 bytes from 172.16.10.1: icmp_seq=3 ttl=255 time=1.44 ms
  7. --- 172.16.10.1 ping statistics ---
  8. 3 packets transmitted, 3 received, 0% packet loss, time 2004ms
  9. rtt min/avg/max/mdev = 1.441/2.163/3.602/1.017 ms
  10. root@server1:~#
  11. root@server1:~# ping -c 3 172.16.20.1 PING 172.16.20.1 (172.16.20.1) 56(84) bytes of data.64 bytes from 172.16.20.1: icmp_seq=1 ttl=255 time=1.35 ms64 bytes from 172.16.20.1: icmp_seq=2 ttl=255 time=1.48 ms64 bytes from 172.16.20.1: icmp_seq=3 ttl=255 time=1.46 ms--- 172.16.20.1 ping statistics ---3 packets transmitted, 3 received, 0% packet loss, time 2004msrtt min/avg/max/mdev = 1.353/1.429/1.477/0.054 msroot@server1:~#
复制代码
网卡ipvlan模式

IPVLAN(IP Virtual LAN)是Linux内核提供的一种网络捏造化技术，它可以在一个物理网卡上创建多个捏造网卡接口，每个捏造接口都有自己独立的IP地点。
IPVLAN和macvlan类似，都是从一个主机接口捏造出多个捏造网络接口。唯一比较大的区别就是ipvlan捏造出的子接口都有类似的mac地点（与物理接口共用同个mac地点），但可配置差别的ip地点。
ipvlan模式下，对端互换机也可以配置access模式或trunk模式，trunk模式下ipvlan可以或许与vlan很好的结合使用。
示意图如下

互换机配置
  1. <H3C>system-view
  2. [H3C]interface GigabitEthernet 1/0/1
  3. [H3C-GigabitEthernet1/0/1]port link-type access
  4. [H3C-GigabitEthernet1/0/1]port access vlan 10
  5. [H3C-GigabitEthernet1/0/1]exit
  6. [H3C]
复制代码
服务器配置，ipvlan支持三种模式(l2、l3、l3s)，这里使用l3模式，并长期化配置
  1. cat >/etc/networkd-dispatcher/routable.d/10-ipvlan-interfaces.sh<<EOF
  2. #! /bin/bash
  3. ip link add ipvlan0 link enp1s0 type ipvlan mode l3
  4. ip link add ipvlan1 link enp1s0 type ipvlan mode l3
  5. EOF
  6. chmod o+x,g+x,u+x /etc/networkd-dispatcher/routable.d/10-ipvlan-interfaces.sh
复制代码
配置netplan
  1. root@server1:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   ethernets:
  4.     enp1s0:
  5.       dhcp4: false
  6.       addresses:
  7.         - 172.16.10.10/24
  8.       nameservers:
  9.         addresses:
  10.           - 223.5.5.5
  11.           - 223.6.6.6
  12.       routes:
  13.         - to: default
  14.           via: 172.16.10.1
  15.     ipvlan0:
  16.       addresses:
  17.         - 172.16.10.11/24
  18.     ipvlan1:
  19.       addresses:
  20.         - 172.16.10.12/24
  21.   version: 2
复制代码
应用网卡配置
  1. netplan apply
复制代码
查看网卡信息，新建了两ipvlan接口，IP地点与主网卡位于同一网段，而且每个接口都有与主网卡类似的MAC地点。
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
  9.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  10.     inet 172.16.10.10/24 brd 172.16.10.255 scope global enp1s0
  11.        valid_lft forever preferred_lft forever
  12.     inet6 fe80::7eb5:9bff:fe59:a71/64 scope link
  13.        valid_lft forever preferred_lft forever
  14. 9: ipvlan0@enp1s0: <BROADCAST,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN group default qlen 1000    link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff    inet 172.16.10.11/24 brd 172.16.10.255 scope global ipvlan0       valid_lft forever preferred_lft forever    inet6 fe80::7cb5:9b00:159:a71/64 scope link        valid_lft forever preferred_lft forever10: ipvlan1@enp1s0: <BROADCAST,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN group default qlen 1000    link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff    inet 172.16.10.12/24 brd 172.16.10.255 scope global ipvlan1       valid_lft forever preferred_lft forever    inet6 fe80::7cb5:9b00:259:a71/64 scope link        valid_lft forever preferred_lft forever
复制代码
测试与网关的连通性
  1. root@server1:~# ping -c 3 172.16.10.1
  2. PING 172.16.10.1 (172.16.10.1) 56(84) bytes of data.
  3. 64 bytes from 172.16.10.1: icmp_seq=1 ttl=255 time=3.60 ms
  4. 64 bytes from 172.16.10.1: icmp_seq=2 ttl=255 time=1.45 ms
  5. 64 bytes from 172.16.10.1: icmp_seq=3 ttl=255 time=1.44 ms
  7. --- 172.16.10.1 ping statistics ---
  8. 3 packets transmitted, 3 received, 0% packet loss, time 2004ms
  9. rtt min/avg/max/mdev = 1.441/2.163/3.602/1.017 ms
  10. root@server1:~#
复制代码
网卡 macvtap 模式

使用 bridge 使 KVM 捏造机可以或许进行外部通信的另一种替代方法是使用 Linux MacVTap 驱动步伐。当不想创建平常网桥，但希望本地网络中的用户访问捏造机时，可以使用 MacVTap。
与使用bridge 的一个主要区别是 MacVTap 直接连接到 KVM 主机中的网络接口。这种直接连接绕过了 KVM 主机中与连接和使用软件bridge 相关的大部门代码和组件，有效地缩短了代码路径。这种较短的代码路径通常会提高吞吐量并减少外部体系的耽误。
示意图如下：

互换机配置
  1. <H3C>system-view
  2. [H3C]interface GigabitEthernet 1/0/1
  3. [H3C-GigabitEthernet1/0/1]port link-type access
  4. [H3C-GigabitEthernet1/0/1]port access vlan 10
  5. [H3C-GigabitEthernet1/0/1]exit
复制代码
主机网卡配置
  1. root@server1:~# cat /etc/netplan/00-installer-config.yaml
  2. # This is the network config written by 'subiquity'
  3. network:
  4.   ethernets:
  5.     enp1s0:
  6.       dhcp4: false
  7.       addresses:
  8.         - 172.16.10.10/24
  9.       nameservers:
  10.         addresses:
  11.           - 223.5.5.5
  12.           - 223.6.6.6
  13.       routes:
  14.         - to: default
  15.           via: 192.168.137.2
  16.   version: 2
复制代码
安装kvm捏造化情况，创建两个捏造机，指定从enp1s0主网卡分配mavtap子接口。
  1. virt-install \
  2.   --name vm1 \
  3.   --vcpus 1 \
  4.   --memory 2048 \
  5.   --disk path=/var/lib/libvirt/images/vm1/jammy-server-cloudimg-amd64.img \
  6.   --os-variant ubuntu22.04 \
  7.   --noautoconsole \
  8.   --import \
  9.   --autostart \
  10.   --network type=direct,source=enp1s0,source_mode=bridge,model=virtio
  12. virt-install \
  13.   --name vm2 \
  14.   --vcpus 1 \
  15.   --memory 2048 \
  16.   --disk path=/var/lib/libvirt/images/vm2/jammy-server-cloudimg-amd64.img \
  17.   --os-variant ubuntu22.04 \
  18.   --noautoconsole \
  19.   --import \
  20.   --autostart \
  21.   --network type=direct,source=enp1s0,source_mode=bridge,model=virtio
复制代码
查看网卡信息，新创建了两个macvtap接口
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
  9.     link/ether 7c:b5:9b:59:0a:71 brd ff:ff:ff:ff:ff:ff
  10.     inet 172.16.10.10/24 brd 172.16.10.255 scope global enp1s0
  11.        valid_lft forever preferred_lft forever
  12.     inet6 fe80::7eb5:9bff:fe59:a71/64 scope link
  13.        valid_lft forever preferred_lft forever
  14. 5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN group default qlen 1000    link/ether 52:54:00:bb:15:22 brd ff:ff:ff:ff:ff:ff    inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0       valid_lft forever preferred_lft forever6: macvtap0@enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 500    link/ether 52:54:00:41:8f:a3 brd ff:ff:ff:ff:ff:ff    inet6 fe80::5054:ff:fe41:8fa3/64 scope link        valid_lft forever preferred_lft forever7: macvtap1@enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 500    link/ether 52:54:00:93:2c:4a brd ff:ff:ff:ff:ff:ff    inet6 fe80::5054:ff:fe93:2c4a/64 scope link        valid_lft forever preferred_lft forever
复制代码
捏造机1配置IP地点
  1. root@vm1:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   ethernets:
  4.     enp1s0:
  5.       dhcp4: false
  6.       addresses:
  7.         - 172.16.10.11/24
  8.       nameservers:
  9.         addresses:
  10.           - 223.5.5.5
  11.           - 223.6.6.6
  12.       routes:
  13.         - to: default
  14.           via: 172.16.10.1
  15.   version: 2
复制代码
捏造机2配置IP地点
  1. root@vm2:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   ethernets:
  4.     enp1s0:
  5.       dhcp4: false
  6.       addresses:
  7.         - 172.16.10.12/24
  8.       nameservers:
  9.         addresses:
  10.           - 223.5.5.5
  11.           - 223.6.6.6
  12.       routes:
  13.         - to: default
  14.           via: 172.16.10.1
  15.   version: 2
复制代码
测试与网关的连通性
  1. root@vm1:~# ping 172.16.10.1 -c 3
  2. PING 172.16.10.1 (172.16.10.1) 56(84) bytes of data.
  3. 64 bytes from 172.16.10.1: icmp_seq=1 ttl=255 time=1.38 ms
  4. 64 bytes from 172.16.10.1: icmp_seq=2 ttl=255 time=1.75 ms
  5. 64 bytes from 172.16.10.1: icmp_seq=3 ttl=255 time=4.34 ms
  7. --- 172.16.10.1 ping statistics ---
  8. 3 packets transmitted, 3 received, 0% packet loss, time 2004ms
  9. rtt min/avg/max/mdev = 1.382/2.491/4.344/1.318 ms
复制代码
bond、vlan与桥接混合配置

将服务器两块网卡构成bond口，在bond口之上创建两个vlan子接口，分别参加两个linux bridge中，然后在差别bridge下创建捏造机，捏造机将属于差别vlan。
示意图如下：

互换机配置，配置动态链路聚合，将端口g1/0/1和g1/0/3参加聚合组。将聚合口配置为trunk模式，允许vlan 8 10 20通过，而且将vlan8 配置为聚合口的native vlan，作为管理使用。
  1. <H3C>system-view
  2. [H3C]interface Vlan-interface 8
  3. [H3C-Vlan-interface8]ip address 172.16.8.1 24
  4. [H3C-Vlan-interface8]exit
  5. [H3C]
  7. [H3C]interface Bridge-Aggregation 1
  8. [H3C-Bridge-Aggregation1]link-aggregation mode dynamic
  9. [H3C-Bridge-Aggregation1]quit
  11. [H3C]interface GigabitEthernet 1/0/1
  12. [H3C-GigabitEthernet1/0/1]port link-aggregation group 1
  13. [H3C-GigabitEthernet1/0/1]exit
  15. [H3C]interface GigabitEthernet 1/0/3
  16. [H3C-GigabitEthernet1/0/3]port link-aggregation group 1
  17. [H3C-GigabitEthernet1/0/3]exit
  19. [H3C]interface Bridge-Aggregation 1
  20. [H3C-Bridge-Aggregation1]port link-type trunk
  21. [H3C-Bridge-Aggregation1]port trunk permit vlan 8 10 20
  22. [H3C-Bridge-Aggregation1]port trunk pvid vlan 8
  23. [H3C-Bridge-Aggregation1]undo port trunk permit vlan 1
  24. [H3C-Bridge-Aggregation1]exit
  25. [H3C]
复制代码
服务器网卡配置，注意bond0配置了管理IP地点，匹配互换机native vlan 8。
  1. root@server1:~# cat /etc/netplan/00-installer-config.yaml
  2. network:
  3.   version: 2
  4.   ethernets:
  5.     enp1s0:
  6.       dhcp4: false
  7.     enp2s0:
  8.       dhcp4: false
  9.   bonds:
  10.     bond0:
  11.       dhcp4: false
  12.       dhcp6: false
  13.       interfaces:
  14.         - enp1s0
  15.         - enp2s0
  16.       addresses:
  17.         - 172.16.8.10/24
  18.       nameservers:
  19.         addresses:
  20.           - 223.5.5.5
  21.           - 223.6.6.6
  22.       routes:
  23.         - to: default
  24.           via: 172.16.8.1
  25.       parameters:
  26.         mode: 802.3ad
  27.         lacp-rate: fast
  28.         mii-monitor-interval: 100
  29.         transmit-hash-policy: layer2+3
  30.   bridges:
  31.     br10:
  32.       interfaces: [ vlan10 ]
  33.     br20:
  34.       interfaces: [ vlan20 ]
  35.   vlans:
  36.     vlan10:
  37.       id: 10
  38.       link: bond0
  39.     vlan20:
  40.       id: 20
  41.       link: bond0
复制代码
查看网卡信息，新建了bond0网口，而且基于bond0网口创建了两个vlan子接口vlan10和vlan20。
  1. root@server1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc fq_codel master bond0 state UP group default qlen 1000
  9.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff permaddr 7c:b5:9b:59:0a:71
  10. 3: enp2s0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc fq_codel master bond0 state UP group default qlen 1000
  11.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff permaddr e4:54:e8:dc:e5:88
  12. 15: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  13.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff
  14.     inet 172.16.8.10/24 brd 172.16.8.255 scope global bond0
  15.        valid_lft forever preferred_lft forever
  16.     inet6 fe80::acfd:60ff:fe48:841a/64 scope link
  17.        valid_lft forever preferred_lft forever
  18. 16: br10: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  19.     link/ether ee:df:66:ab:c2:4b brd ff:ff:ff:ff:ff:ff
  20.     inet6 fe80::ecdf:66ff:feab:c24b/64 scope link
  21.        valid_lft forever preferred_lft forever
  22. 17: br20: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
  23.     link/ether 9e:4d:f4:0a:6d:13 brd ff:ff:ff:ff:ff:ff
  24.     inet6 fe80::9c4d:f4ff:fe0a:6d13/64 scope link
  25.        valid_lft forever preferred_lft forever
  26. 18: vlan10@bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master br10 state UP group default qlen 1000
  27.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff
  28. 19: vlan20@bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master br20 state UP group default qlen 1000
  29.     link/ether ae:fd:60:48:84:1a brd ff:ff:ff:ff:ff:ff
复制代码
查看创建的网桥
  1. root@server1:~# brctl show
  2. bridge name     bridge id               STP enabled     interfaces
  3. br10            8000.eedf66abc24b       no              vlan10
  4. br20            8000.9e4df40a6d13       no              vlan20
复制代码
测试bond0 IP与外部网关连通性
  1. root@server1:~# ping 172.16.8.1 -c 3
  2. PING 172.16.8.1 (172.16.8.1) 56(84) bytes of data.
  3. 64 bytes from 172.16.8.1: icmp_seq=1 ttl=255 time=1.55 ms
  4. 64 bytes from 172.16.8.1: icmp_seq=2 ttl=255 time=1.61 ms
  5. 64 bytes from 172.16.8.1: icmp_seq=3 ttl=255 time=1.62 ms
  7. --- 172.16.8.1 ping statistics ---
  8. 3 packets transmitted, 3 received, 0% packet loss, time 2004ms
  9. rtt min/avg/max/mdev = 1.550/1.593/1.620/0.030 ms
  10. root@server1:~#
复制代码
在server1安装kvm捏造化情况，然后创建两个新的kvm网络，分别绑定到差别网桥
  1. cat >br10-network.xml<<EOF
  2. <network>
  3.   <name>br10-net</name>
  4.   <forward mode="bridge"/>
  5.   <bridge name="br10"/>
  6. </network>
  7. EOF
  8. cat >br20-network.xml<<EOF
  9. <network>
  10.   <name>br20-net</name>
  11.   <forward mode="bridge"/>
  12.   <bridge name="br20"/>
  13. </network>
  14. EOF
  16. virsh net-define br10-network.xml
  17. virsh net-define br20-network.xml
  18. virsh net-start br10-net
  19. virsh net-start br20-net
  20. virsh net-autostart br10-net
  21. virsh net-autostart br20-net
复制代码
查看新建的网络
  1. root@server1:~# virsh net-list
  2. Name       State    Autostart   Persistent
  3. ---------------------------------------------
  4. br10-net   active   yes         yes
  5. br20-net   active   yes         yes
  6. default    active   yes         yes
复制代码
创建两个捏造机，指定使用差别网络
  1. virt-install \
  2.   --name vm1 \
  3.   --vcpus 1 \
  4.   --memory 2048 \
  5.   --disk path=/var/lib/libvirt/images/vm1/jammy-server-cloudimg-amd64.img \
  6.   --os-variant ubuntu22.04 \
  7.   --import \
  8.   --autostart \
  9.   --noautoconsole \
  10.   --network network=br10-net
  12. virt-install \
  13.   --name vm2 \
  14.   --vcpus 1 \
  15.   --memory 2048 \
  16.   --disk path=/var/lib/libvirt/images/vm2/jammy-server-cloudimg-amd64.img \
  17.   --os-variant ubuntu22.04 \
  18.   --import \
  19.   --autostart \
  20.   --noautoconsole \
  21.   --network network=br20-net
复制代码
查看创建的捏造机
  1. root@server1:~# virsh list
  2. Id   Name   State
  3. ----------------------
  4. 13   vm1    running
  5. 14   vm2    running
复制代码
为vm1配置vlan10的IP地点
  1. virsh console vm1
  2. cat >/etc/netplan/00-installer-config.yaml<<EOF
  3. network:
  4.   ethernets:
  5.     enp1s0:
  6.       addresses:
  7.       - 172.16.10.10/24
  8.       nameservers:
  9.         addresses:
  10.         - 223.5.5.5
  11.       routes:
  12.       - to: default
  13.         via: 172.16.10.1
  14.   version: 2
  15. EOF
  16. netplan apply
复制代码
为vm2配置vlan20的IP地点
  1. virsh console vm2
  2. cat >/etc/netplan/00-installer-config.yaml<<EOF
  3. network:
  4.   ethernets:
  5.     enp1s0:
  6.       addresses:
  7.       - 172.16.20.10/24
  8.       nameservers:
  9.         addresses:
  10.         - 223.5.5.5
  11.       routes:
  12.       - to: default
  13.         via: 172.16.20.1
  14.   version: 2
  15. EOF
  16. netplan apply
复制代码
登录到vm1，测试vm1与外部网关连通性
  1. root@vm1:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
  9.     link/ether 52:54:00:a4:aa:9d brd ff:ff:ff:ff:ff:ff
  10.     inet 172.16.10.10/24 brd 172.16.10.255 scope global enp1s0
  11.        valid_lft forever preferred_lft forever
  12.     inet6 fe80::5054:ff:fea4:aa9d/64 scope link
  13.        valid_lft forever preferred_lft forever
  14. root@vm1:~#
  15. root@vm1:~# ping 172.16.10.1 -c 3
  16. PING 172.16.10.1 (172.16.10.1) 56(84) bytes of data.
  17. 64 bytes from 172.16.10.1: icmp_seq=1 ttl=255 time=1.51 ms
  18. 64 bytes from 172.16.10.1: icmp_seq=2 ttl=255 time=7.10 ms
  19. 64 bytes from 172.16.10.1: icmp_seq=3 ttl=255 time=2.10 ms
  21. --- 172.16.10.1 ping statistics ---
  22. 3 packets transmitted, 3 received, 0% packet loss, time 2003ms
  23. rtt min/avg/max/mdev = 1.505/3.568/7.101/2.509 ms
  24. root@vm1:~#
复制代码
登录到vm2，测试vm2与外部网关连通性
  1. root@vm2:~# ip a
  2. 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
  3.     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  4.     inet 127.0.0.1/8 scope host lo
  5.        valid_lft forever preferred_lft forever
  6.     inet6 ::1/128 scope host
  7.        valid_lft forever preferred_lft forever
  8. 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
  9.     link/ether 52:54:00:89:61:da brd ff:ff:ff:ff:ff:ff
  10.     inet 172.16.20.10/24 brd 172.16.20.255 scope global enp1s0
  11.        valid_lft forever preferred_lft forever
  12.     inet6 fe80::5054:ff:fe89:61da/64 scope link
  13.        valid_lft forever preferred_lft forever
  14. root@vm2:~#
  15. root@vm2:~# ping 172.16.20.1 -c 3
  16. PING 172.16.20.1 (172.16.20.1) 56(84) bytes of data.
  17. 64 bytes from 172.16.20.1: icmp_seq=1 ttl=255 time=1.73 ms
  18. 64 bytes from 172.16.20.1: icmp_seq=2 ttl=255 time=2.00 ms
  19. 64 bytes from 172.16.20.1: icmp_seq=3 ttl=255 time=2.00 ms
  21. --- 172.16.20.1 ping statistics ---
  22. 3 packets transmitted, 3 received, 0% packet loss, time 2003ms
  23. rtt min/avg/max/mdev = 1.732/1.911/2.003/0.126 ms
  24. root@vm2:~#
复制代码
免责声明：如果侵犯了您的权益，请联系站长，我们会及时删除侵权内容，谢谢合作！更多信息从访问主页：qidao123.com:ToB企服之家，中国第一个企服评测及商务社交产业平台。

本帖子中包含更多资源

您需要 登录 才可以下载或查看，没有账号？立即注册

x
回复

使用道具 举报

0 个回复

倒序浏览
返回列表

快速回复

高级模式
B Color Image Link Quote Code Smilies
您需要登录后才可以回帖 登录 or 立即注册

本版积分规则

天空闲话

金牌会员
这个人很懒什么都没写!

楼主热帖

标签云

挺好的 服务器
微信订阅号
微信服务号
微信客服
小程序
H5
关于我们 商务合作 网站地图
©Copyright 2022-2024 杭州祥升科技有限公司 版权所有 浙ICP备20004199号
快速回复 返回顶部 返回列表