0 Replies Latest reply on Nov 12, 2018 2:58 PM by bwatty

    No Reply to Ping When Not Originating from VMnet8 NAT Subnet

    bwatty Lurker

      I am using VMWare Workstation 11 and I'm not receiving replies to pings when I am at least one subnet removed from the vmnet8 NAT subnet. I have the following configuration:

       

      Host: Windows 7

      Guest VMs: Red Hat Linux

       

      Wireless Access Point IP Address: 192.168.170.1

      Wireless Network LAN connection on host pc: 192.168.170.73

       

      Server 1 - vmnet10 subnet: 10.0.1.0/24

                       eth1 ip address: 10.0.1.1

       

                       default gw: 10.0.1.200

       

      Server 2 - vmnet10 subnet: 10.0.1.0/24

                       eth0 ip address: 10.0.1.200

       

                       vmnet8 NAT subnet: 192.168.25.0/24

                       eth1 ip address: 192.168.25.128

       

                       default gw: 192.168.25.2

       

      If I ping from Server 2 (192.168.25.128) to the NAT device/gateway at 192.168.25.2 I get replies. I also get replies if I ping the Wireless Network LAN connection (192.168.170.73)  as well as the Wireless Access Point ip address (192.168.170.1). The arp requests from the NAT device want to know who-has 192.168.25.128 and are given the correct response with the MAC address.

       

      If I ping from Server 1 (10.0.1.1) to the same ip addresses there is no reply or response to the ping. In tcpdump I see that arp requests originate from the NAT device (192.168.25.2) and it is querying to see who-has the source ip address (10.0.1.1). Since the source ip address didn't originate from vmnet8 as 192.168.25.128 then it is met with no response. From a routing perspective it would seem that vmnet8 would only need to know how to get packets back to 192.168.25.128 as a gateway to the other networks.

       

      It seems as if the NAT device doesn't have access to the routing tables where it would be able to send arp requests to the proper ip addresses for routing purposes.

       

      Has this been a known issue and are there any workarounds?