I just started using OpenVPN cloud connectors for creating mesh VPN networks. I deployed a connector and setup relvant IP forwarding and NAT protocols that allow me to access the LAN network behind the machine with runs the OpenVPN cloud connector. I was able to do that successfully. (my info is mostly from this link https://openvpn.net/cloud-docs/connecting-networks-to-openvpn-cloud-using-connectors/)
Now the problem is why am I not able to SSH into the machine that runs the cloud connector over its VPN IP of 100.96.1.18? Weirdly, I can ping it no problem.
~$ ssh admin@100.96.1.18
ssh: connect to host 100.96.1.18 port 22: Network is unreachable
~$ ping 100.96.1.18 -c 4
PING 100.96.1.18 (100.96.1.18) 56(84) bytes of data.
64 bytes from 100.96.1.18: icmp_seq=1 ttl=62 time=209 ms
64 bytes from 100.96.1.18: icmp_seq=2 ttl=62 time=279 ms
64 bytes from 100.96.1.18: icmp_seq=3 ttl=62 time=208 ms
64 bytes from 100.96.1.18: icmp_seq=4 ttl=62 time=204 ms
--- 100.96.1.18 ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3004ms
rtt min/avg/max/mdev = 203.702/224.971/279.214/31.383 ms
Surprisingly, I can remotely (through a VPN client machine) SSH into the connector machine over its Local IP of 192.168.18.1
Here are the IP routes on the connector machine.
~$ ip route
default via 192.168.18.1 dev eno1 proto dhcp metric 100
64.120.110.199 via 192.168.18.1 dev eno1
100.80.0.0/12 via 100.96.1.17 dev tun0
100.96.0.0/11 via 100.96.1.17 dev tun0
100.96.1.16/30 dev tun0 proto kernel scope link src 100.96.1.18
169.254.0.0/16 dev eno1 scope link metric 1000
172.17.0.0/16 dev docker0 proto kernel scope link src 172.17.0.1
192.168.0.0/24 via 100.96.1.17 dev tun0
192.168.18.0/24 dev eno1 proto kernel scope link src 192.168.18.88 metric 100
~$ ifconfig tun0
tun0: flags=4305<UP,POINTOPOINT,RUNNING,NOARP,MULTICAST> mtu 1500
inet 100.96.1.18 netmask 255.255.255.252 destination 100.96.1.18
inet6 fd:0:0:8101::2 prefixlen 126 scopeid 0x0<global>
inet6 fe80::5a84:3a5:f59b:d64f prefixlen 64 scopeid 0x20<link>
unspec 00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00 txqueuelen 500 (UNSPEC)
RX packets 1682 bytes 451719 (451.7 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1755 bytes 637526 (637.5 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
Also for debug here are the IP routes from a connected client machine
~$ ip route
default via 192.168.31.1 dev wlp2s0 proto dhcp metric 600
25.0.0.0/8 dev ham0 proto kernel scope link src 25.56.7.62
100.80.0.0/12 via 100.96.1.1 dev tun0
100.96.0.0/11 via 100.96.1.1 dev tun0
100.96.1.0/28 dev tun0 proto kernel scope link src 100.96.1.2
169.254.0.0/16 dev wlp2s0 scope link metric 1000
172.17.0.0/16 dev docker0 proto kernel scope link src 172.17.0.1 linkdown
172.107.213.76 via 192.168.31.1 dev wlp2s0
192.168.0.0/24 via 100.96.1.1 dev tun0
192.168.18.0/24 via 100.96.1.1 dev tun0
192.168.31.0/24 dev wlp2s0 proto kernel scope link src 192.168.31.189 metric 600
~$ traceroute 100.96.1.18
traceroute to 100.96.1.18 (100.96.1.18), 64 hops max
1 100.96.1.18 265.200ms !N 109.653ms !N 105.821ms !N
My main point of contention is that since 100.96.0.0/11 via 100.96.1.1 dev tun0 route is being pushed to the client I should be able to reach the 100.96.1.18 IP and ssh or traceroute etc. to it.
P.S. Asking this here since OpenVPN support couldn't answer.
