Challenge w/ network routing - had to set lots of routes to see remote host

[srekrap]

Hello -

  So I have two hosts, TC (tinycore) and RH (RHEL6). Both have two nics, each host on two same subnets, say "10." and "158."

  Both TC and RH hosts have the same default gateway route to a "158..129" address, which both can ping.

  On the RH host I can ping a "far off" host, call it hostX (..254.18 below), and when I traceroute hostX I get 6 or so hops through specific IP's, the gateway 158 host being the first.

  [root@RH shared2]# traceroute 158.187.254.18
  traceroute to 158.187.254.18 (158.187.254.18), 30 hops max, 60 byte packets
   1  158.187.254.129 (158.187.254.129)  0.473 ms  0.640 ms  0.750 ms
   2  158.187.216.1 (158.187.216.1)  0.422 ms  0.500 ms  0.524 ms
   3  158.187.245.105 (158.187.245.105)  0.365 ms  0.448 ms  0.484 ms
   4  158.187.245.110 (158.187.245.110)  0.363 ms  0.628 ms  0.675 ms
   5  158.187.188.12 (158.187.188.12)  0.426 ms  0.544 ms  0.625 ms
   6  158.187.254.18 (158.187.254.18)  0.319 ms  0.302 ms  0.290 ms
  [root@RH shared2]#

  On the TC host I can ping RH host, and gatewaye host,  and many other "close" hosts that can also be pinged by the RH host.

  However, the only way I can ping hostX (158.187.254.18 ) from the TC host is if I add a route for each and EVERY hopped-host (from RH traceroute run above) using the gateway host (158.187.254.129) as the gateway.

  Am I doing something whacked here? (I don't have any ip table stuff configured on TC or RH hosts, and I don't have access to any of the other hosts as they are enterprise controlled.) I've used the IP for the TC host on other RH distros in our env and never had this problem, so I don't the network is doing anything to block this IP in particular...

  Any help appreciated...

   Thanks, Sean

[srekrap]

... as a follow-up, here's the route table from the TC host:

(This is what I had to do in order to be able to ping 158.187.254.18)


tc@TC:~$ route
Kernel IP routing table
Destination        Gateway                 Genmask         Flags Metric Ref    Use Iface
default               158.187.254.129   0.0.0.0         UG    0      0        0 eth2
10.0.0.0              *                           255.0.0.0       U     0      0        0 eth0
127.0.0.1            *                           255.255.255.255 UH    0      0        0 lo
158.187.0.0        *                           255.255.0.0     U     0      0        0 eth2
158.187.188.12   158.187.254.129 255.255.255.255 UGH   0      0        0 eth2
158.187.216.1      158.187.254.129 255.255.255.255 UGH   0      0        0 eth2
158.187.245.105 158.187.254.129 255.255.255.255 UGH   0      0        0 eth2
158.187.245.110 158.187.254.129 255.255.255.255 UGH   0      0        0 eth2
158.187.254.18   158.187.254.129 255.255.255.255 UGH   0      0        0 eth2
tc@TC:~$ 


This is route on RH host:

[root@gateway6 shared2]# route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
192.168.122.0   *               255.255.255.0   U     0      0        0 virbr0
10.128.0.0      *               255.255.0.0     U     0      0        0 eth1
link-local      *               255.255.0.0     U     1002   0        0 eth0
link-local      *               255.255.0.0     U     1003   0        0 eth1
default         158.187.254.129 0.0.0.0         UG    0      0        0 eth0
[root@gateway6 shared2]#

[srekrap]

If I remove the preceeding routes and leave the route to .18, I can still ping .18, but didn't think I'd need to add .18  as a route...

[curaga]

You didn't post the traceroute to .18 on the TC box, without any extra routes.

If the traceroute reaches your gateway, but no further, that gateway is dropping your packet.

[tinypoodle]

How come you are using Lockheed Martin as your default gateway?   

[gerald_clark]

route add default gw  158.187.254.129

[srekrap]

@tinypoodle - 'cause that's where I work!

@gerald_clark - yep - that's how I got the route entries in both TC and RH - I'm confused about why I need to add a specific route to .18 when I do have the defaults in there... that's my concern with this issue

@curaga - I didn't post 'cause it didn't make it past the .129 host (before adding route to .18) - if that GW is dropping my packats, then I guess I don't know why or how-to/have-rights to fix it... for now I'll just add the route to the .18 box -

If no one sees anything obvious then I can deal with it... just an annoyance

Thanks everyone for your time!

[gerald_clark]

You cannot use dhcp for two interfaces.
You need to manually set up your network in /opt/bootlocal.sh ( or /opt/bootsync.sh ).

[tinypoodle]

@tinypoodle - 'cause that's where I work!

That still doesn't explain why you would use a public IP as default gateway.

A traceroute leading through 3 different subnets only to end at a destination IP in same subnet as source IP is rather odd...

Difficult to say more if you do not provide more details about your network topology.
Potentially "hairpinning" might play a role.

[srekrap]

@gerald_clark - these are manual configurations for testing, and no DHCP in effect here - all static IPs.

@tinypoodle - I'm in a lab, and have no control over the topology. I don't know what's public or private - if you think the hairpinning might be a problem, I'll ask around to see. Thanks!