Client 장비에 network port가 여러개 있는 경우, 특정 network port를 지정하여 IP 패킷을 전송하고 싶을 때가 있다.
이럴 때, source IP address를 binding하면 특정 network port를 통해 IP 패킷이 전송된다.
참고: 일반적으로 Target IP address로 가기 위한 routing path 및 network port는 OS에 있는 Routing table을 통해서 자동으로 결정된다. 그러나 Target IP address로 가기 위한 routing path가 1개가 아닌 2개 이상인 경우에는 어느 network port를 통해서 IP 패킷이 나갈지 예측하기 어렵다.
package main
import (
"fmt"
"io/ioutil"
"net"
"net/http"
"time"
)
func main() {
##
## NOTE: 14.33.80.179를 Source IP address로 지정한다. (즉, Source IP address binding)
##
localAddr, err := net.ResolveIPAddr("ip", "14.33.80.179")
if err != nil {
panic(err)
}
localTCPAddr := net.TCPAddr{
IP: localAddr.IP,
}
d := net.Dialer{
LocalAddr: &localTCPAddr,
Timeout: 30 * time.Second,
KeepAlive: 30 * time.Second,
}
tr := &http.Transport{
Proxy: http.ProxyFromEnvironment,
Dial: d.Dial,
TLSHandshakeTimeout: 10 * time.Second,
}
webclient := &http.Client{Transport: tr}
// Use NewRequest so we can change the UserAgent string in the header
req, err := http.NewRequest("GET", "https://www.naver.com", nil)
if err != nil {
panic(err)
}
res, err := webclient.Do(req)
if err != nil {
panic(err)
}
fmt.Println("DEBUG", res)
defer res.Body.Close()
content, err := ioutil.ReadAll(res.Body)
if err != nil {
panic(err)
}
fmt.Printf("%s", string(content))
}
South Carolina 대학의 'Open Virtual Switch Lab Series' 문서를 바탕으로 내가 실습한 내용을 이곳에 정리함. ( Network namespace 개념부터 차곡차곡 쌓아 올리면서 Open vSwitch Use Case를 설명하기 때문에 공부하는 사람에게 많은 도움이 된다 )
위 그림에 묘사된 것과 같이 Network를 구성하기 위해 아래 명령을 작성했다. (따라해보면 위 그림과 똑같은 Network 만들어진다)
## root namespace에 존재하는 모든 network interface를 조회
$ ip link
## 네임스페이스 my-ns-a, my-ns-b 를 생성
$ ip netns add my-ns-a
$ ip netns add my-ns-b
## Linux kernel에 존재하는 모든 namespace 조회
$ ip netns
my-ns-b
my-ns-a
## 'my-ns-a' 네임스페이스에 존재하는 network interface 조회
$ ip netns exec my-ns-a ip link
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
## 가상 스위치 'sw1'를 생성
$ ovs-vsctl add-br sw1
## root namespace에 존재하는 network interface를 조회
$ ip link
... 중간 생략 ...
47: ovs-system: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 86:3d:02:69:23:4f brd ff:ff:ff:ff:ff:ff
48: sw1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 16:68:07:5d:c0:40 brd ff:ff:ff:ff:ff:ff
## Open vSwitch에 namespace를 연결하기
## 1) veth peer 생성하기
$ ip link add my-ns-a-eth0 type veth peer name sw1-eth1
$ ip link add my-ns-b-eth0 type veth peer name sw1-eth2
$ ip link
... 중간 생략 ...
47: ovs-system: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 86:3d:02:69:23:4f brd ff:ff:ff:ff:ff:ff
48: sw1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 16:68:07:5d:c0:40 brd ff:ff:ff:ff:ff:ff
51: sw1-eth1@my-ns-a-eth0: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether be:01:52:6f:4b:58 brd ff:ff:ff:ff:ff:ff
52: my-ns-a-eth0@sw1-eth1: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 96:24:a4:bf:78:f3 brd ff:ff:ff:ff:ff:ff
53: sw1-eth2@my-ns-b-eth0: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 46:d4:ad:57:18:20 brd ff:ff:ff:ff:ff:ff
54: my-ns-b-eth0@sw1-eth2: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 2a:78:4d:57:db:37 brd ff:ff:ff:ff:ff:ff
## 2) veth peer를 각각의 namepace에 연결하기 (Attaching to namespaces)
$ ip link set my-ns-a-eth0 netns my-ns-a
$ ip link set my-ns-b-eth0 netns my-ns-b
$ ip netns exec my-ns-a ip link
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
52: my-ns-a-eth0@if51: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 96:24:a4:bf:78:f3 brd ff:ff:ff:ff:ff:ff link-netnsid 0
$ ip netns exec my-ns-b ip link
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
54: my-ns-b-eth0@if53: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 2a:78:4d:57:db:37 brd ff:ff:ff:ff:ff:ff link-netnsid 0
## 3) 가상 스위치 sw1에 veth peer를 연결하기 (Attaching veth peer to switch sw1)
$ ovs-vsctl add-port sw1 sw1-eth1
$ ovs-vsctl show
...
Bridge sw1
Port sw1
Interface sw1
type: internal
Port sw1-eth1
Interface sw1-eth1
...
$ ovs-vsctl add-port sw1 sw1-eth2
$ ovs-vsctl show
...
Bridge sw1
Port sw1
Interface sw1
type: internal
Port sw1-eth2
Interface sw1-eth2
Port sw1-eth1
Interface sw1-eth1
...
## 가상 스위치의 network port를 activate 하기. (Turning up the network port)
$ ip link set sw1-eth1 up
$ ip link set sw1-eth2 up
$ ip link
...
51: sw1-eth1@if52: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue master ovs-system state LOWERLAYERDOWN mode DEFAULT group default qlen 1000
link/ether be:01:52:6f:4b:58 brd ff:ff:ff:ff:ff:ff link-netns my-ns-a
53: sw1-eth2@if54: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue master ovs-system state LOWERLAYERDOWN mode DEFAULT group default qlen 1000
link/ether 46:d4:ad:57:18:20 brd ff:ff:ff:ff:ff:ff link-netns my-ns-b
...
## 각각의 namespace에 IP address를 할당하기
$ ip netns exec my-ns-a ip link set dev my-ns-a-eth0 up
$ ip netns exec my-ns-b ip link set dev my-ns-b-eth0 up
$ ip netns exec my-ns-a ip address add 192.168.1.10/24 dev my-ns-a-eth0
$ ip netns exec my-ns-b ip address add 192.168.1.20/24 dev my-ns-b-eth0
## 설정 정보 확인하기
$ ip netns exec my-ns-a ip addr
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
52: my-ns-a-eth0@if51: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 96:24:a4:bf:78:f3 brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet 192.168.1.10/24 scope global my-ns-a-eth0
valid_lft forever preferred_lft forever
inet6 fe80::9424:a4ff:febf:78f3/64 scope link
valid_lft forever preferred_lft forever
$ ip netns exec my-ns-b ip addr
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
54: my-ns-b-eth0@if53: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 2a:78:4d:57:db:37 brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet 192.168.1.20/24 scope global my-ns-b-eth0
valid_lft forever preferred_lft forever
inet6 fe80::2878:4dff:fe57:db37/64 scope link
valid_lft forever preferred_lft forever
## namespace 'my-ns-a'의 routing table 확인하기
$ ip netns exec my-ns-a ip route
192.168.1.0/24 dev my-ns-a-eth0 proto kernel scope link src 192.168.1.10
## namespace 'my-ns-b'의 routing table 확인하기
$ ip netns exec my-ns-b ip route
192.168.1.0/24 dev my-ns-b-eth0 proto kernel scope link src 192.168.1.20
## namespace 'my-ns-a'에서 bash shell 시작하기
$ ip netns exec my-ns-a bash
$ ifconfig
my-ns-a-eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.10 netmask 255.255.255.0 broadcast 0.0.0.0
inet6 fe80::9424:a4ff:febf:78f3 prefixlen 64 scopeid 0x20<link>
ether 96:24:a4:bf:78:f3 txqueuelen 1000 (Ethernet)
RX packets 86 bytes 21517 (21.5 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 13 bytes 1006 (1.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
$ ping 192.168.1.20 -c 2
PING 192.168.1.20 (192.168.1.20) 56(84) bytes of data.
64 bytes from 192.168.1.20: icmp_seq=1 ttl=64 time=0.088 ms
64 bytes from 192.168.1.20: icmp_seq=2 ttl=64 time=0.079 ms
--- 192.168.1.20 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1007ms
rtt min/avg/max/mdev = 0.079/0.083/0.088/0.004 ms
$ traceroute 192.168.1.20
traceroute to 192.168.1.20 (192.168.1.20), 64 hops max
1 192.168.1.20 0.452ms 0.003ms 0.002ms
