我有 2 个 Linux 盒 A 和 B,都是 RHEL 7。
B网卡配置:
eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 9001
inet 10.74.61.131 netmask 255.255.254.0 broadcast 10.74.61.255
inet6 fe80::1011:12ff:fec7:d118 prefixlen 64 scopeid 0x20<link>
ether 12:11:12:c7:d1:18 txqueuelen 1000 (Ethernet)
RX packets 4599 bytes 503250 (491.4 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 4489 bytes 693330 (677.0 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
eth1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 9001
inet 10.74.60.97 netmask 255.255.254.0 broadcast 10.74.61.255
inet6 fe80::10b3:c9ff:fe85:47b5 prefixlen 64 scopeid 0x20<link>
ether 12:b3:c9:85:47:b5 txqueuelen 1000 (Ethernet)
RX packets 136 bytes 8864 (8.6 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 274 bytes 22184 (21.6 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
B 内核 IP 路由表
Destination Gateway Genmask Flags Metric Ref Use Iface
0.0.0.0 10.74.60.1 0.0.0.0 UG 100 0 0 eth0
0.0.0.0 10.74.60.1 0.0.0.0 UG 101 0 0 eth1
10.74.60.0 0.0.0.0 255.255.254.0 U 100 0 0 eth0
10.74.60.0 0.0.0.0 255.255.254.0 U 101 0 0 eth1
146.1.242.0 10.74.61.26 255.255.254.0 UG 0 0 0 eth1
B ip 路由
default via 10.74.60.1 dev eth0 proto static metric 100
default via 10.74.60.1 dev eth1 proto static metric 101
10.74.60.0/23 dev eth0 proto kernel scope link src 10.74.61.131 metric 100
10.74.60.0/23 dev eth1 proto kernel scope link src 10.74.60.97 metric 101
146.1.242.0/23 via 10.74.61.26 dev eth1
B路由到146.1.242.62:
ip route get 146.1.242.62
146.1.242.62 via 10.74.61.26 dev eth1 src 10.74.60.97
cache
网卡配置
eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 9001
inet 10.74.61.26 netmask 255.255.254.0 broadcast 10.74.61.255
ether 12:d5:c1:64:c1:84 txqueuelen 1000 (Ethernet)
RX packets 41439 bytes 2210425 (2.1 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 78272 bytes 23367775 (22.2 MiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
eth1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 9001
inet 10.74.60.101 netmask 255.255.254.0 broadcast 10.74.61.255
ether 12:f6:c9:b0:75:05 txqueuelen 1000 (Ethernet)
RX packets 521 bytes 32470 (31.7 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 275 bytes 22494 (21.9 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
内核 IP 路由表
Destination Gateway Genmask Flags Metric Ref Use Iface
0.0.0.0 10.74.60.1 0.0.0.0 UG 100 0 0 eth0
0.0.0.0 10.74.60.1 0.0.0.0 UG 101 0 0 eth1
10.74.60.0 0.0.0.0 255.255.254.0 U 100 0 0 eth0
10.74.60.0 0.0.0.0 255.255.254.0 U 101 0 0 eth1
IP 路由
default via 10.74.60.1 dev eth0 proto static metric 100
default via 10.74.60.1 dev eth1 proto static metric 101
10.74.60.0/23 dev eth0 proto kernel scope link src 10.74.61.26 metric 100
10.74.60.0/23 dev eth1 proto kernel scope link src 10.74.60.101 metric 101
A 用作网关,在 B 中,任何发往 146.1.242.0/23 的流量都会通过具有 10.74.61.26 网关的 eth1。
B --> eth1(gw 10.74.61.26)----->A(eth0 10.74.61.26)
到 146.1.242.62 的路由:
ip route get 146.1.242.62
146.1.242.62 via 10.74.61.26 dev eth1 src 10.74.60.97
cache
我已经制作了可路由的 IP 路由配置 /etc/sysctl.conf
net.ipv4.ip_forward = 1
net.ipv6.conf.eth0.disable_ipv6 = 1
#disable ipv6
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1
net.ipv6.conf.lo.disable_ipv6 = 1
net.ipv6.conf.eth0.disable_ipv6 = 1
net.ipv4.conf.default.rp_filter = 2
net.ipv4.conf.all.rp_filter = 2
已启用并启动防火墙。Iptables 全部接受,没有丢弃或拒绝任何内容。
因此当我从 B ping 时
ping -I eth1 146.1.242.62
PING 146.1.242.62 (146.1.242.62) from 10.74.60.97 eth1: 56(84) bytes of data.
^C
--- 146.1.242.62 ping statistics ---
2 packets transmitted, 0 received, 100% packet loss, time 999ms
它从来没有通过,我做错了什么。
当我从 B ping 到 A 时,以下是 tcpdump 结果
ping -I eth0 146.1.242.62
PING 146.1.242.62 (146.1.242.62) from 10.74.61.131 eth0: 56(84) bytes of data.
64 bytes from 146.1.242.62: icmp_seq=1 ttl=50 time=36.8 ms
^C
--- 146.1.242.62 ping statistics ---
6 packets transmitted, 6 received, 0% packet loss, time 5009ms
tcpdump icmp 响应:
IP 146.1.242.62 > ip-10-74-61-131.ebiz.ee.com: ICMP echo reply, id 3405, seq 1, length 64
IP ip-10-74-61-131.ebiz.ee.com > 146.1.242.62: ICMP echo request, id 3405, seq 2, length 64
IP 146.1.242.62 > ip-10-74-61-131.ebiz.ee.com: ICMP echo reply, id 3405, seq 2, length 64
IP ip-10-74-61-131.ebiz.ee.com > 146.1.242.62: ICMP echo request, id 3405, seq 3, length 64
IP 146.1.242.62 > ip-10-74-61-131.ebiz.ee.com: ICMP echo reply, id 3405, seq 3, length 64
IP ip-10-74-61-131.ebiz.ee.com > 146.1.242.62: ICMP echo request, id 3405, seq 4, length 64
IP 146.1.242.62 > ip-10-74-61-131.ebiz.ee.com: ICMP echo reply, id 3405, seq 4, length 64
IP ip-10-74-61-131.ebiz.ee.com > 146.1.242.62: ICMP echo request, id 3405, seq 5, length 64
但当我这样做
ping -I eth1 146.1.242.62
PING 146.1.242.62 (146.1.242.62) from 10.74.60.97 eth1: 56(84) bytes of data.
^C
--- 146.1.242.62 ping statistics ---
4 packets transmitted, 0 received, 100% packet loss, time 2999ms
没有 tcpdump icmp 响应。
截至 2016 年 11 月 21 日 A 的 iptables
iptables -t nat -S
-P PREROUTING ACCEPT
-P INPUT ACCEPT
-P OUTPUT ACCEPT
-P POSTROUTING ACCEPT
...
-A PREROUTING -j LOG --log-prefix "-- PRE ROUTE--" --log-level 6
-A POSTROUTING -j LOG --log-prefix "-- POST ROUTE--" --log-level 6
iptables -S
-P INPUT ACCEPT
-P FORWARD ACCEPT
-P OUTPUT ACCEPT
B的iptables:
iptables -t nat -S
-P PREROUTING ACCEPT
-P INPUT ACCEPT
-P OUTPUT ACCEPT
-P POSTROUTING ACCEPT
-A POSTROUTING -j LOG --log-prefix "-- POST ROUTE--" --log-level 6
在 B 中启用后路由日志并在 A 中启用预路由和后路由日志后,我 ping 了目标 (146.1.242.62),它在 B 后路由中记录,但在 A 的预路由中没有任何记录。以下是日志条目和 ping
ping 146.1.242.62
PING 146.1.242.62 (146.1.242.62) from 10.74.60.97 eth1: 56(84) bytes of data.
^C
--- 146.1.242.62 ping statistics ---
4 packets transmitted, 0 received, 100% packet loss, time 2999ms
B log:
Nov 21 10:36:25 localhost kernel: -- POST ROUTE--IN= OUT=eth1 SRC=10.74.60.97 DST=146.1.242.62 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=19877 DF PROTO=ICMP TYPE=8 CODE=0 ID=4920 SEQ=1
A log: none, no log in PRE ROUTE & POST ROUTE about the ping
答案1
您的帖子未显示任何 iptables 规则。我猜想在您的主机 A 上,您应该确保通过的流量经过 NAT。
以 root 身份运行以下命令可能会有帮助:
iptables -t nat -A POSTROUTING --src 10.74.60.97/32 \! --dst 10.74.61.26/32 -o eth0 -j MASQUERADE
您可能还会发现tcpdump
有用的,可以解决网络问题。