CentOS 双网卡绑定实现平衡负载

绑定两块网卡主要为了解决网卡故障、负载均衡等问题。

1、在vm加一块网卡，登录后检查网卡是否识别。

分别用ip addr和nmcli查看网卡的情况
[[email protected]-senior01 ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:ea:31:47 brd ff:ff:ff:ff:ff:ff
    inet 192.168.31.10/24 brd 192.168.31.255 scope global ens33
       valid_lft forever preferred_lft forever
    inet6 fe80::20c:29ff:feea:3147/64 scope link
       valid_lft forever preferred_lft forever
3: ens37: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:ea:31:51 brd ff:ff:ff:ff:ff:ff
新加入的网卡接口是ens37

2、常见的绑定模式

常用的有三种
mode=0：平衡负载模式，有自动备援，但需要交换机支持及设定，       两个交换机端口需要做聚合，该模式下bond所绑定的网卡的IP都被修改成相同的mac地址，交换机做了聚合后，聚合下的几个端口也被捆绑成一个mac地址。
mode=1：自动备援模式，其中一条线若断线，其他线路将会自动备援。
mode=6：平衡负载模式，有自动备援，不用交换机支持，绑定的是不同的MAC地址。

3、用例，环境CentOS7.x（CentOS6是另外的配置）

#在/etc/sysconfig/network-scripts目录下配置ens33,ens37,bond0三个网卡接口,bond0是抽象网卡
[[email protected]-senior01 network-scripts]# cat ifcfg-ens33
TYPE=Ethernet
BOOTPROTO=none
DEVICE=ens33
ONBOOT=yes
USERCTL=no
MASTER=bond0
SLAVE=yesBONDING_MASTER=yes

[[email protected]-senior01 network-scripts]# cat ifcfg-ens37
TYPE=Ethernet
BOOTPROTO=none
DEVICE=ens37
ONBOOT=yes
USERCTL=no
MASTER=bond0
SLAVE=yesBONDING_MASTER=yes
#实际是把原来网卡1的ip配置放入了bond0里
BOOTPROTO=noneDEVICE=bond0TYPE=BondONBOOT=yesZONE=publicIPADDR=192.168.31.10NETMASK=255.255.255.0GATEWAY=192.168.31.2DNS1=192.168.31.2USERCTL=noNM_CONTROLLED=noBONDING_MASTER=yesBONDING_OPTS="mode=6 miimon=100"

重启网络systemctl restart network，如果原来network就没启动，那么使用systemctl start network

#ens33和ens37应该是没有配置IP的。[[email protected] modprobe.d]# ip addr1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00    inet 127.0.0.1/8 scope host lo       valid_lft forever preferred_lft forever    inet6 ::1/128 scope host        valid_lft forever preferred_lft forever2: ens33: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 state UP qlen 1000    link/ether 00:0c:29:ea:31:47 brd ff:ff:ff:ff:ff:ff3: ens37: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 state UP qlen 1000    link/ether 00:0c:29:ea:31:47 brd ff:ff:ff:ff:ff:ff4: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP qlen 1000    link/ether 00:0c:29:ea:31:47 brd ff:ff:ff:ff:ff:ff    inet 192.168.31.10/24 brd 192.168.31.255 scope global bond0       valid_lft forever preferred_lft forever    inet6 fe80::20c:29ff:feea:3147/64 scope link tentative dadfailed        valid_lft forever preferred_lft forever

查看bonding的状态[[email protected] ~]# cat /proc/net/bonding/bond0 Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)

Bonding Mode: adaptive load balancingPrimary Slave: NoneCurrently Active Slave: ens33MII Status: upMII Polling Interval (ms): 100Up Delay (ms): 0Down Delay (ms): 0

Slave Interface: ens33MII Status: upSpeed: 1000 MbpsDuplex: fullLink Failure Count: 0Permanent HW addr: 00:0c:29:ea:31:47Slave queue ID: 0

Slave Interface: ens37MII Status: upSpeed: 1000 MbpsDuplex: fullLink Failure Count: 0Permanent HW addr: 00:0c:29:ea:31:51Slave queue ID: 0

#测试，在ping的过程中中断一个网卡的连接（可以在VM的属性里取消连接勾选），丢包2%，网卡从ens33自动切成ens34--- 192.168.1.103 ping statistics ---86 packets transmitted, 84 received, 2% packet loss, time 85275msrtt min/avg/max/mdev = 0.471/1.057/1.684/0.348 ms

原文地址：https://www.cnblogs.com/asker009/p/10290128.html