[TriLUG] some TCP utilities ignore kernel routing table?

[Brian Henning]

Here's a bit more info on the layout here.

We have two physical /24 subnets in the building: 192.168.1.0 and 
192.168.10.0.  They're physically separate; switched by separate 
switchgear and all that good stuff.  .125 has a NIC on both subnets 
(.1.125 and .10.125), and is configured to act as a router between the two.

My workstation, .1.32, is only on the .1 subnet and has a routing table 
entry informing it how to reach .10.  Previously we had a host on the 
.10 subnet with a complementary routing table entry telling it how to 
reach .1, and we had communication between the subnets with no difficulty.

The target now, .10.4, is a Gestetner P7145 with onboard 
IPP/TCP/JetDirect/etc. network print services.  While it can't handle 
special routing, it does have a default gateway setting, which I have 
set to .10.125, and this appears to work inasmuch as I can ping it from 
.1.32 as well as reach its http config interface from .1.32.

But, as shown previously, telnet can't find it, traceroute reports some 
errors, and cups won't talk to it.

~Brian

Ron Joffe wrote:
> On Tuesday 10 January 2006 12:38, Brian Henning wrote:
> 
>>Hi Folks,
>>   Just wondering if someone could shed some light on why it seems that
>>some tcp utilities seem to ignore the kernel routing table.  For
>>example:  (Sorry about the wrapping..)
> 
> 
> Could the problem be on the remote end. The packets might be getting there, 
> but might not know how to route back to your box.
> 
> Is this a vpn or tunnel that you have set up with forwarding from 1.124 to the 
> 192.168.10.0 network?
> 
> Ron
> 

-- 
----------------
Brian A. Henning
strutmasters.com
336.597.2397x238
----------------