0

Apologies for the terrible title, it's the best I could do.

This is my network:

schematic

  • Modem is just a plain dumb wired ADSL modem

  • Router 1 is a wireless router / access point (TP-Link WDR-4300 with latest official firmware, to be specific), with a number of clients, both wired and wireless, connected to it (even though the picture only shows two). It's configured with a static route to send all traffic headed for 192.168.2.0/24 to router 2.

  • Router 2 is a wireless router (WRT54GL with DD-WRT firmware) configured to act as a wireless client, connected to router 1

The problem: 192.168.2.10, connected to router 2, cannot access any of the clients connected to router 1 - pings do not receive responses, ssh / smbclient will time out. It has access to internet with no issues, and has no problem pinging / telnet'ing any of the routers along the way. Running wireshark / tcpdump on 192.168.0.101 and 192.168.0.102 shows that pings from 192.168.2.10 are definitely getting to them (with the correct source address), so it seems that the replies are just not being received at the source.

192.168.2.10 is definitely accessible from all clients connected to router 1 - they can ping it, connect to smb shares and copy files left and right, ssh into it, you name it.

It looks like the only packets getting lost are those that are replies to something originating 192.168.2.10, which makes little sense to me. What could be wrong here?

  • I think packets are not correctly routed to the 192.168.0.0/24 network. Client 192.168.2.10 is not aware of the other network. – Aulis Ronkainen Aug 18 '18 at 9:26
  • Can you clarify why router 1 has two addresses from the same subnet, and router 2 only has one address at all? Since routers connect networks together, they normally have an address for each different network they're on (i.e. at least two total). – grawity Aug 18 '18 at 12:32
  • The green boxes represent clients, not the addresses of the routers (not a great diagram, and I only included a subset of the clients for show). Router 1 has internal IP 192.168.0.1 for a bridge that combines the wireless and ethernet switch, and gets its WAN IP from the modem by DHCP. Router 2 has internal IP 192.168.2.1 for the wired side, and wireless IP 192.168.0.2 (since it is connected to router 1 as a wireless client). – jaymmer Aug 18 '18 at 13:16
  • I am still confused about your description. Is Router 2 connected to Router 1 with a wired connection while being a wireless client of Router 1? Have you tried to connect directly to the LAN port of Router 2 with the LAN port of Router 1? – S.Leon Aug 20 '18 at 9:56
  • From your observations/tests, I don't think this is a routing issue. So issue should lie in a firewall configuration along the way (either on the routers or on the client). When you say "Running wireshark / tcpdump on 192.168.0.101 and 192.168.0.102 shows that pings from 192.168.2.10 are definitely getting to them" => Do you see ICMP response packets leaving 192.168.0.101-102 ? – Gohu Aug 22 '18 at 12:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.