Timeout pings from router to clients

RouterOS General
1

Hi to everyone,

I’ve recently done some big changes in my network. I moved from one big bridged scenario with 6 distribution nodes, to a routed scenario with 8 different network segments. Everything is working seamlessly from clients point of view (better than before), but it is not the same for me.

My main structure is the following:
Clients (10.20.10.0/24) —))) Node_i <-(10.20.7.0/24)–> UM_Gateway <–192.168.0.0/29–> RB532_LoadBalancing <-192.168.254.0/29-> line1, line2 and a Management Computer to winbox everything and manage usermanager.

From any client in any node I can ping/traceroute to everything (internet or my network, including RB532 and Computer) and get internet connection.

But, I can only ping/traceroute/access my clients from its nearest node and from UM_Gateway. I can’t reach them from RB532 or from Computer.

I can ping/traceroute/access node_i clients’ side Address (10.20.10.1, for instance) from RB532 and Computer. So this problem only appears when node_i forwards traffic.

If I try to traceroute a working client (10.20.10.189 for example) from RB532 or from computer, trace gets stopped at last hop.

[admin@rb532_loadbalancing] > /tool traceroute 10.20.10.189
     ADDRESS                                    STATUS
   1 192.168.0.2     3ms 2ms 1ms 
   2 10.20.7.6       3ms 2ms 2ms 
   3 0.0.0.0         timeout timeout timeout 
   4 0.0.0.0         timeout timeout timeout 
   5 0.0.0.0         timeout timeout timeout 
[b]Here it stays timing out forever[/b]
[admin@rb532_loadbalancing] >

Hop 3 should be 10.20.10.189.

  • Node_i default GW: UM_Gateway
  • UM_Gateway default gw: RB532_LoadBalancing and static routes to each node_i (10.20.10.0/23, gw Node_i)
  • RB532_LoadBalancing default gw: Line1. Static route to 10.20.0.0/16 gw: UM_Gateway. Static route to Line2.
  • Computer default gw: RB532_LoadBalancing. Reach Internet through Line1. Has direct ethernet connection to 10.20.7.0/24 but no route to client networks using this interface.
  • Doesn’t matter which node is node_i, I have the same behaviour on all nodes.

I have double and triple checked there’s no firewall rule or queue that could affect my traffic.

Anyone could see an small ray of light or has any small idea about?

2

You’ve only routed one way - from client to headend.

You know have to route from headend to client. 1 hop at a time.