聚合链路与桥接测试

Posted 2020-10-09 遠離塵世の方舟

用途：将主机上的两块网卡绑定在一起，构成一个逻辑端口。
交换机根据用户配置的端口负荷分担策略决定报文从哪一个成员端口发送到对端的交换机。当其中一块网卡发生故障时，则继续使用另外一块正常的网卡。

RHEL5/6叫网卡bonding，需要加载内核模块
RHEL7叫链路聚合，内核自带驱动，需要配置team的master和多个slave

聚合网络的原理

真实的ip地址并不是在物理网卡上设置的，而是把两个或多个物理网卡聚合成一个虚拟的网卡，在虚拟网卡上设置地址，而外部网络访问本机时，就是访问的这个虚拟网卡的地址，虚拟网卡接到数据后经过两个网卡的负载交给服务器处理。如果一块网卡出现问题，则通过另一块传递数据，保证正常通信。聚合网络实验过程中，添加的两块新网卡可以不是active状态，甚至nmcli connect show查看没有新添加的网卡信息。也可以使active状态的网卡。

链路聚合的工作模式
1.active-backup  主备
2.loadbalance    负载均衡
3.RR  RoundRobin    轮询

测试环境：

[root@localhost ~]# ifconfig enp0s3: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500          inet 192.168.100.4  netmask 255.255.255.255  broadcast 192.168.100.4          inet6 fe80::922e:b8be:7ea8:4562  prefixlen 64  scopeid 0x20<link>          ether 08:00:27:3e:f6:4d  txqueuelen 1000  (Ethernet)          RX packets 164  bytes 20454 (19.9 KiB)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 168  bytes 28639 (27.9 KiB)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536          inet 127.0.0.1  netmask 255.0.0.0          inet6 ::1  prefixlen 128  scopeid 0x10<host>          loop  txqueuelen 1  (Local Loopback)          RX packets 82  bytes 8488 (8.2 KiB)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 82  bytes 8488 (8.2 KiB)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 virbr0: flags=4099<UP,BROADCAST,MULTICAST>  mtu 1500          inet 192.168.122.1  netmask 255.255.255.0  broadcast 192.168.122.255          ether 52:54:00:81:e7:fc  txqueuelen 1000  (Ethernet)          RX packets 0  bytes 0 (0.0 B)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 0  bytes 0 (0.0 B)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 [root@localhost ~]# hostnamectl     Static hostname: localhost.localdomain           Icon name: computer-vm             Chassis: vm          Machine ID: c59cb0b9011f459297a11b7b6b3e457b             Boot ID: b029771ca9204611abe911f960ea7d51      Virtualization: kvm    Operating System: Red Hat Enterprise Linux Server 7.4 (Maipo)         CPE OS Name: cpe:/o:redhat:enterprise_linux:7.4:GA:server              Kernel: Linux 3.10.0-693.el7.x86_64        Architecture: x86-64 [root@localhost ~]# nmcli connection show NAME    UUID                                  TYPE            DEVICE enp0s3  59118063-4e4e-403c-ad25-374704d6f6f1  802-3-ethernet  enp0s3 virbr0  b7180c81-0027-49ee-ab06-8f34a2df30be  bridge          virbr0 [root@localhost ~]#

一、测试主备模式（active-backup）

1、在主机模式再添加一块新的网卡

先shutdown 虚拟机，然后再 VirtualBox界面再启用网卡2。

然后启动虚拟机。

2、查看网卡信息

[root@localhost ~]# nmcli device show | grep -i device --新增网卡为enp0s8 GENERAL.DEVICE:                         virbr0 GENERAL.DEVICE:                         enp0s3 GENERAL.DEVICE:                         enp0s8 GENERAL.DEVICE:                         lo GENERAL.DEVICE:                         virbr0-nic [root@localhost ~]# nmcli connection show ---看不到enp0s8，因为无配置文件 NAME    UUID                                  TYPE            DEVICE enp0s3  59118063-4e4e-403c-ad25-374704d6f6f1  802-3-ethernet  enp0s3 virbr0  b4f4c345-d863-41c0-b396-55fd0f67785f  bridge          virbr0

# nmcli connection show 看不到enp0s8，因为无配置文件，如果手动添加配置文件后，就可以看到enp0s8了。其实都没必要为enp0s8和enp0s3造配置文件，因为我们要跟把两块网卡绑定到一块，作为新的team设备，使用新的配置文件。这里只为了测试一个小知识点。

[root@localhost ~]# nmcli connection add con-name enp0s8 type ethernet ifname enp0s8 Connection \'enp0s8\' (b3972183-4a71-4635-b910-02c0530517d3) successfully added. [root@localhost ~]# nmcli connection show NAME    UUID                                  TYPE            DEVICE enp0s3  59118063-4e4e-403c-ad25-374704d6f6f1  802-3-ethernet  enp0s3 enp0s8  b3972183-4a71-4635-b910-02c0530517d3  802-3-ethernet  enp0s8 virbr0  b4f4c345-d863-41c0-b396-55fd0f67785f  bridge          virbr0

3、添加team设备及配置文件

[root@localhost ~]# nmcli connection add type team con-name testteamfile ifname testteamdevice config \'{"runner":{"name":"activebackup"}}\' Connection \'testteamfile\' (d2c8f925-8645-4ac5-95e6-6245a4431778) successfully added. [root@localhost ~]# nmcli connection show --如果看不到team设备，说明上一命令有问题。 NAME          UUID                                  TYPE            DEVICE         enp0s3        59118063-4e4e-403c-ad25-374704d6f6f1  802-3-ethernet  enp0s3         enp0s8        b3972183-4a71-4635-b910-02c0530517d3  802-3-ethernet  enp0s8         testteamfile  d2c8f925-8645-4ac5-95e6-6245a4431778  team            testteamdevice virbr0        b4f4c345-d863-41c0-b396-55fd0f67785f  bridge          virbr0

说明：

# nmcli connection add type team con-name testteamfile ifname testteamdevice config \'{"runner":{"name":"activebackup"}}\'
testteamfile配置文件名
testteamdevice 绑定的设备名
config ’{运行规则}’。运行规则中出现的所有关键字使用双引号。
"runner":{"name":"activebackup"}  指明运行级别为activebackup

如果添加错误，则使用 nmcli connection delete testteamfile来删除，然后重新再添加。

配置文件添加成功后，会在/etc/sysconfig/network-scripts/目录下生成。

[root@localhost ~]# cd /etc/sysconfig/network-scripts/
[root@localhost network-scripts]# ls ifcfg-*
ifcfg-enp0s3  ifcfg-enp0s8  ifcfg-lo  ifcfg-testteamfile

4、为team设备配置IP

[root@localhost network-scripts]# nmcli connection modify testteamfile ipv4.method manual ipv4.addresses 192.168.100.123/24 ipv4.gateway 192.168.100.100 ipv4.dns 192.168.100.1 connection.autoconnect yes [root@localhost network-scripts]# nmcli connection show NAME          UUID                                  TYPE            DEVICE         enp0s3        59118063-4e4e-403c-ad25-374704d6f6f1  802-3-ethernet  enp0s3         enp0s8        b3972183-4a71-4635-b910-02c0530517d3  802-3-ethernet  enp0s8         testteamfile  d2c8f925-8645-4ac5-95e6-6245a4431778  team            testteamdevice virbr0        b4f4c345-d863-41c0-b396-55fd0f67785f  bridge          virbr0

5、将两块网卡enp0s3和enp0s8绑定到虚拟网卡上

[root@localhost ~]# nmcli connection add type team-slave con-name testteamslave1 ifname enp0s3 master testteamdevice Connection \'testteamslave1\' (506b5fe9-0c3b-4251-b58b-b1a1a53acda4) successfully added. [root@localhost ~]# nmcli connection add type team-slave con-name testteamslave2 ifname enp0s8 master testteamdevice Connection \'testteamslave2\' (312ba2d8-7b18-4446-b37f-5de7217a6646) successfully added. [root@localhost ~]# nmcli connection show NAME            UUID                                  TYPE            DEVICE enp0s3          59118063-4e4e-403c-ad25-374704d6f6f1  802-3-ethernet  enp0s3 enp0s8          b3972183-4a71-4635-b910-02c0530517d3  802-3-ethernet  enp0s8 virbr0          b4f4c345-d863-41c0-b396-55fd0f67785f  bridge          virbr0 testteamfile    d2c8f925-8645-4ac5-95e6-6245a4431778  team            --     testteamslave1  506b5fe9-0c3b-4251-b58b-b1a1a53acda4  802-3-ethernet  --     testteamslave2  312ba2d8-7b18-4446-b37f-5de7217a6646  802-3-ethernet  --

enp0s3----->testteamslave1------>testteamdevice

enpos8----->testteamslave2------>testteamdevice

可以这么理解：enp0s3和enp0s8被它们的主子master买去做了奴隶，一个取名叫slave1，另一个取名叫slave2。

6、激活虚拟网卡testteamdevice

[root@localhost ~]# nmcli connection up testteamfile  ---注意启动的是文件，而不是设备 Connection successfully activated (master waiting for slaves) (D-Bus active path: /org/freedesktop/NetworkManager/ActiveConnection/157) [root@localhost ~]# nmcli connection up testteamslave1 [root@localhost ~]# nmcli connection up testteamslave2

先启动master,再启动两个slave。

[root@localhost ~]# nmcli connection show NAME            UUID                                  TYPE            DEVICE         virbr0-nic      6944d05c-3906-4fc5-84e7-cb46f3e6e2b2  generic         virbr0-nic     virbr0          05e6ca14-f6f5-4697-bcc0-e13e4b212236  bridge          virbr0         enp0s3          7419645c-73a9-423f-bcfd-fccd36773841  802-3-ethernet  --             testteamslave2  608bb6f1-099d-459d-9589-5b2f2b28b2d9  802-3-ethernet  enp0s8         testteamslave1  90b0d08f-dec2-4a1f-827a-ce198f21fb46  802-3-ethernet  enp0s3         testteamfile    f2277aa4-a7b3-4166-afaa-dcd60e550d97  team            testteamdevice [root@localhost ~]# ifconfig | grep ether          ether 08:00:27:57:c1:ff  txqueuelen 1000  (Ethernet)          ether 08:00:27:57:c1:ff  txqueuelen 1000  (Ethernet)          ether 08:00:27:57:c1:ff  txqueuelen 0  (Ethernet)          ether 52:54:00:59:75:9b  txqueuelen 0  (Ethernet) [root@localhost ~]# ifconfig enp0s3: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500          ether 08:00:27:57:c1:ff  txqueuelen 1000  (Ethernet)          RX packets 144  bytes 17362 (16.9 KiB)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 152  bytes 23282 (22.7 KiB)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 enp0s8: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500          ether 08:00:27:57:c1:ff  txqueuelen 1000  (Ethernet)          RX packets 10  bytes 907 (907.0 B)          RX errors 0  dropped 5  overruns 0  frame 0          TX packets 0  bytes 0 (0.0 B)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536          inet 127.0.0.1  netmask 255.0.0.0          inet6 ::1  prefixlen 128  scopeid 0x10<host>          loop  txqueuelen 0  (Local Loopback)          RX packets 10  bytes 984 (984.0 B)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 10  bytes 984 (984.0 B)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 testteamdevice: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500          inet 192.168.100.123  netmask 255.255.255.0  broadcast 192.168.100.255          inet6 fe80::a00:27ff:fe57:c1ff  prefixlen 64  scopeid 0x20<link>          ether 08:00:27:57:c1:ff  txqueuelen 0  (Ethernet)          RX packets 133  bytes 14204 (13.8 KiB)          RX errors 0  dropped 4  overruns 0  frame 0          TX packets 116  bytes 17841 (17.4 KiB)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 virbr0: flags=4099<UP,BROADCAST,MULTICAST>  mtu 1500          inet 192.168.122.1  netmask 255.255.255.0  broadcast 192.168.122.255          ether 52:54:00:59:75:9b  txqueuelen 0  (Ethernet)          RX packets 0  bytes 0 (0.0 B)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 0  bytes 0 (0.0 B)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 You have mail in /var/spool/mail/root

从上边可以看出:testteamdevice,enp0s3,enp0s8 三块网卡的mac地址一样。

三个常用的命令：

1、查看虚拟网卡的状态

[root@localhost ~]# teamdctl testteamdevice state setup:    runner: activebackup   --运行模式 ports:    enp0s3      link watches:        link summary: up        instance[link_watch_0]:          name: ethtool          link: up    --状态为启动          down count: 0    enp0s8      link watches:        link summary: up        instance[link_watch_0]:          name: ethtool          link: up   --状态为启动          down count: 0 runner:    active port: enp0s3 --当前激活的网卡；testteamdevice设备目前使用的是enp0s3网卡

2、查看虚拟网卡的端口

[root@localhost ~]# teamnl testdevice -h teamnl [options] teamdevname command [command args]          -h --help                Show this help Commands:          ports          options          getoption OPT_NAME          setoption OPT_NAME OPT_VALUE          monitor OPT_STYLE [root@localhost ~]# teamnl testteamdevice ports    ----查看网卡对应的端口   3: enp0s8: up 1000Mbit FD   2: enp0s3: up 1000Mbit FD [root@localhost ~]# teamnl testteamdevice getoption activeport  ---当前使用的网卡端口，当前使用的是enp0s3。 2

验证结果：

[root@localhost ~]# nmcli device disconnect enp0s3   --断开enp0s3 Device \'enp0s3\' successfully disconnected. [root@localhost ~]# teamdctl testteamdevice state ---可以虚拟网卡自动连接到时enp0s8网卡上 [root@localhost ~]#teamncl testteamdevice port 3：enp0s8:up 1000Mbit FD

图形化界面配置方式：

#nm-connection-editor

同样，再添加team设备的另外一块网卡进来。

为team2设备配置IP地址：

 

二、桥接

要求：把本机1的其中一块网卡转换为桥接设备（交换机），其它客户机连接到该设备后能与主机1正常通信。

下面把网卡enp0s8转换为桥接设备：

[root@localhost ~]# nmcli connection show NAME    UUID                                  TYPE            DEVICE enp0s3  6759482b-076f-4181-9d69-ca6dc0190faa  802-3-ethernet  enp0s3 virbr0  4f99dff9-d3fb-461a-b0d6-00297e7584a0  bridge          virbr0 [root@localhost ~]# nmcli device show  | grep -i device GENERAL.DEVICE:                         virbr0 GENERAL.DEVICE:                         enp0s3 GENERAL.DEVICE:                         enp0s8 GENERAL.DEVICE:                         lo GENERAL.DEVICE:                         virbr0-nic [root@localhost ~]# ifconfig enp0s3: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500          inet 192.168.100.111  netmask 255.255.255.0  broadcast 192.168.100.255          inet6 fe80::6ea:59df:4b18:530b  prefixlen 64  scopeid 0x20<link>          ether 08:00:27:3e:f6:4d  txqueuelen 1000  (Ethernet)          RX packets 188  bytes 21984 (21.4 KiB)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 246  bytes 38495 (37.5 KiB)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 enp0s8: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500          ether 08:00:27:45:59:30  txqueuelen 1000  (Ethernet)          RX packets 63  bytes 9404 (9.1 KiB)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 0  bytes 0 (0.0 B)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536          inet 127.0.0.1  netmask 255.0.0.0          inet6 ::1  prefixlen 128  scopeid 0x10<host>          loop  txqueuelen 1  (Local Loopback)          RX packets 84  bytes 9340 (9.1 KiB)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 84  bytes 9340 (9.1 KiB)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 virbr0: flags=4099<UP,BROADCAST,MULTICAST>  mtu 1500          inet 192.168.122.1  netmask 255.255.255.0  broadcast 192.168.122.255          ether 52:54:00:81:e7:fc  txqueuelen 1000  (Ethernet)          RX packets 0  bytes 0 (0.0 B)          RX errors 0  dropped 0  overruns 0  frame 0          TX packets 0  bytes 0 (0.0 B)          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0 [root@localhost ~]#

#nmcli connection add type bridge con-name testbrifile ifname testbridevice
#nmcli connection modify testbrifile ipv4.method manual ipv4.addresses "192.168.100.222/24 192.168.100.1" ipv4.dns 192.168.100.1
#nmcli connection add type bridge-slave con-name testbrislave1 ifname enpos8 master testbridevice
#nmcli connection down testbrifile
#nmcli connection up testbrifile
#nmcli connection up testbrislave1
#ping -I testbridevice 192.168.100.111    #-I表示从哪块网卡发出ping