Ethernet level failure between hosts? Self-hosted setup

So I have a cloud-hosted planet server that is also joined to a private network. It can talk to other computers, except at one site. At that one site I have an older computer that was already joined, and it can connect but all new ones I try to use fail.

Host shows unreachable:

% ping 172.27.24.40                                                                          22:29:03
PING 172.27.24.40 (172.27.24.40) 56(84) bytes of data.
From 172.27.22.22 icmp_seq=1 Destination Host Unreachable
From 172.27.22.22 icmp_seq=2 Destination Host Unreachable
From 172.27.22.22 icmp_seq=3 Destination Host Unreachable

And here’s the tcpdump showing ARP going unanswered:

% sudo tcpdump -ni ztaua4bbbf host 172.27.22.22                                              22:29:28
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on ztaua4bbbf, link-type EN10MB (Ethernet), capture size 262144 bytes
22:30:04.318234 ARP, Request who-has 172.27.24.40 tell 172.27.22.22, length 28
22:30:05.346019 ARP, Request who-has 172.27.24.40 tell 172.27.22.22, length 28
22:30:06.370000 ARP, Request who-has 172.27.24.40 tell 172.27.22.22, length 28
22:30:06.834622 IP 172.27.22.22.33649 > 239.255.255.250.1900: UDP, length 101
22:30:07.002084 IP 172.27.22.22.47824 > 172.27.127.255.32412: UDP, length 21
22:30:07.002100 IP 172.27.22.22.58727 > 172.27.127.255.32414: UDP, length 21
22:30:07.394207 ARP, Request who-has 172.27.24.40 tell 172.27.22.22, length 28
22:30:08.418030 ARP, Request who-has 172.27.24.40 tell 172.27.22.22, length 28
22:30:09.442027 ARP, Request who-has 172.27.24.40 tell 172.27.22.22, length 28
22:30:10.466174 ARP, Request who-has 172.27.24.40 tell 172.27.22.22, length 28
22:30:11.490016 ARP, Request who-has 172.27.24.40 tell 172.27.22.22, length 28
22:30:12.003395 IP 172.27.22.22.58727 > 172.27.127.255.32414: UDP, length 21
22:30:12.003411 IP 172.27.22.22.47824 > 172.27.127.255.32412: UDP, length 21

The computers eventually join the network (I’m guessing the old computer relays for them? It takes upwards of 5 minutes…), but services are unavailable, either between the new computers themselves, or any others at that site as well…

Does anyone have any idea how to further troubleshoot this? Thanks in advance for the help…

P.S. - I have tried rebooting the network gear, the computers themselves, working with a combination of windows and linux server equipment. The site is on one single NAT’d LAN, so everything comes from the same public IP if that matters?

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.