一：不给eth1分配IP（ipvlan+whereabouts的方案）
二：给eth1分配IP，但用ipvl子接口禁用eth1的数据转发(terway方案）
二与一的优势对比：
二使得同一node下的pod可以与node直接通讯而不需要走vpc网关。
一中所有pod与node的通讯都需要走vpc网关（从eth1->gw->eth0）。
terway ipvlan子接口依赖的主网卡的使用上的细节:
一旦主网卡定义声明为ipvlan L2 模式，那么这张网卡就无法向外发送流量
在node上的作用依旧是正常的：
- 主网卡的ip依然可以ping通
- 可以作为“网桥” 让所有slave子接口互通
所以目前使用的双网卡模式，在eth1未初始化ipvlan模式之前，仍具有转发功能，但是eth0负责提供负载均衡器依赖的vip。
eth0 eth1分属于两个子网，但属于同一个vpc，连接到同一个软路由。为了保证 vip对应的ipvs后端pod的流量走eth0，在初始化ipvlan cni 子接口之前必须关闭eth1，防止由eth1转发pod的流量。
延伸：
当eth1已创建了ipvlan 子接口，此时down掉eth1有什么影响
关闭eth1后：
- ipvlan pod无法ping通网关
- 本地网桥的功能依然正常，同一node上的pod依然可以互通。
- eth1 无法向外转发（和up的时候效果是一样的）
[root@(l2)k8s-master-1 ~]# grep eth1 /etc/cni/net.d/10-terway.conf && ip link set eth1 down
"master": "eth1"
[root@(l2)k8s-master-1 ~]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
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: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1400 qdisc fq_codel state UP group default qlen 1000
link/ether 00:00:00:83:c5:17 brd ff:ff:ff:ff:ff:ff
inet 172.32.16.11/16 brd 172.32.255.255 scope global dynamic noprefixroute eth0
valid_lft 85132245sec preferred_lft 85132245sec
inet6 fe80::200:ff:fe83:c517/64 scope link
valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST> mtu 1400 qdisc fq_codel state DOWN group default qlen 1000
link/ether 00:00:00:b2:8f:1b brd ff:ff:ff:ff:ff:ff
4: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc fq_codel state DOWN group default qlen 1000
link/ether 00:00:00:9c:1c:c7 brd ff:ff:ff:ff:ff:ff
5: eth3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1400 qdisc fq_codel state UP group default qlen 1000
link/ether 00:00:00:90:05:f0 brd ff:ff:ff:ff:ff:ff
inet 10.5.208.208/24 brd 10.5.208.255 scope global dynamic noprefixroute eth3
valid_lft 85132245sec preferred_lft 85132245sec
inet6 fe80::47ae:2181:3ed0:2fdb/64 scope link noprefixroute
valid_lft forever preferred_lft forever
6: docker0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN group default
link/ether 02:42:67:72:ec:1a brd ff:ff:ff:ff:ff:ff
inet 172.17.0.1/16 brd 172.17.255.255 scope global docker0
valid_lft forever preferred_lft forever
7: kube-ipvs0: <BROADCAST,NOARP> mtu 1500 qdisc noop state DOWN group default
link/ether 4e:49:bf:d2:ab:fc brd ff:ff:ff:ff:ff:ff
inet 10.96.0.1/32 scope global kube-ipvs0
valid_lft forever preferred_lft forever
inet 10.96.0.3/32 scope global kube-ipvs0
valid_lft forever preferred_lft forever
inet 10.97.3.203/32 scope global kube-ipvs0
valid_lft forever preferred_lft forever
inet 10.96.165.187/32 scope global kube-ipvs0
valid_lft forever preferred_lft forever
9: ipvl_3@eth1: <BROADCAST,MULTICAST,NOARP,UP,LOWER_UP,M-DOWN> mtu 1400 qdisc noqueue state UNKNOWN group default
link/ether 00:00:00:b2:8f:1b brd ff:ff:ff:ff:ff:ff
inet 172.33.192.10/32 scope host ipvl_3
valid_lft forever preferred_lft forever
inet6 fe80::2b2:8f1b/64 scope link
valid_lft forever preferred_lft forever
[root@(l2)k8s-master-1 ~]#
[root@(l2)k8s-master-1 ~]#
[root@(l2)k8s-master-1 ~]# kubectl get pod -A -o wide | grep k8s-master-1
default busybox1056-k8s-master-1 1/1 Running 0 12d 172.33.0.218 k8s-master-1 <none> <none>
default centos-sshd-cluster-41 1/1 Running 0 6d17h 172.33.0.201 k8s-master-1 <none> <none>
default centos-sshd-cluster-43 1/1 Running 0 6d17h 172.33.0.208 k8s-master-1 <none> <none>
default sshd-k8s-master-1 1/1 Running 0 10d 172.33.0.211 k8s-master-1 <none> <none>
kube-system kube-apiserver-k8s-master-1 1/1 Running 6 14d 172.32.16.11 k8s-master-1 <none> <none>
kube-system kube-controller-manager-k8s-master-1 1/1 Running 11 14d 172.32.16.11 k8s-master-1 <none> <none>
kube-system kube-ovn-cni-zzqpz 1/1 Running 3 13d 172.32.16.11 k8s-master-1 <none> <none>
kube-system kube-proxy-dbbgv 1/1 Running 4 14d 172.32.16.11 k8s-master-1 <none> <none>
kube-system kube-scheduler-k8s-master-1 1/1 Running 12 14d 172.32.16.11 k8s-master-1 <none> <none>
[root@(l2)k8s-master-1 ~]# ping 172.33.0.218
PING 172.33.0.218 (172.33.0.218) 56(84) bytes of data.
64 bytes from 172.33.0.218: icmp_seq=1 ttl=64 time=0.106 ms
64 bytes from 172.33.0.218: icmp_seq=2 ttl=64 time=0.072 ms
64 bytes from 172.33.0.218: icmp_seq=3 ttl=64 time=0.070 ms
64 bytes from 172.33.0.218: icmp_seq=4 ttl=64 time=0.082 ms
64 bytes from 172.33.0.218: icmp_seq=5 ttl=64 time=0.057 ms
^C
--- 172.33.0.218 ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4108ms
rtt min/avg/max/mdev = 0.057/0.077/0.106/0.018 ms
[root@(l2)k8s-master-1 ~]#
[root@(l2)k8s-master-1 ~]#
[root@(l2)k8s-master-1 ~]#
[root@(l2)k8s-master-1 ~]# kubectl exec -ti sshd-k8s-master-1 -- /bin/sh
sh-4.2# ping 114.114.114.114
PING 114.114.114.114 (114.114.114.114) 56(84) bytes of data.
^C
--- 114.114.114.114 ping statistics ---
2 packets transmitted, 0 received, 100% packet loss, time 1052ms
sh-4.2# route -n
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
0.0.0.0 172.33.0.1 0.0.0.0 UG 0 0 0 eth0
172.33.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth0
172.33.192.10 0.0.0.0 255.255.255.255 UH 0 0 0 eth0
sh-4.2# ping 172.33.0.1
PING 172.33.0.1 (172.33.0.1) 56(84) bytes of data.
^C
--- 172.33.0.1 ping statistics ---
2 packets transmitted, 0 received, 100% packet loss, time 1008ms
sh-4.2# ping 172.33.192.10
PING 172.33.192.10 (172.33.192.10) 56(84) bytes of data.
64 bytes from 172.33.192.10: icmp_seq=1 ttl=64 time=0.063 ms
64 bytes from 172.33.192.10: icmp_seq=2 ttl=64 time=0.126 ms
^C
--- 172.33.192.10 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1048ms
rtt min/avg/max/mdev = 0.063/0.094/0.126/0.032 ms
sh-4.2# ping 172.33.0.218
PING 172.33.0.218 (172.33.0.218) 56(84) bytes of data.
64 bytes from 172.33.0.218: icmp_seq=1 ttl=64 time=0.058 ms
64 bytes from 172.33.0.218: icmp_seq=2 ttl=64 time=0.153 ms
^C
--- 172.33.0.218 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1056ms
rtt min/avg/max/mdev = 0.058/0.105/0.153/0.048 ms
网友评论