can't ping single host

[rkmorrow]

This is a revamp of the previous post, i guess I left some information that would make this a little less obvious to resolve.

I have a win2k laptop computer on a subnet trying to reach an ibm mainframe. The computer has connected to the mainframe in the past.There are 30 other people on the same subnet as the laptop connecting to the mainframe. The laptop has the same TCP/IP configuration as the other people in the building that have ip connectivty to the mainframe.

The clients go to a hub with a 1600 router then via point-to-point line to a 2600 router to a switch. The mainframe is connected to the switch.There are no access lists or filtering preventing anyone from connecting to the mainframe. The connection is as follows:


laptop---1600----2600---switch---mainframe

The laptop has suddenly lost ip connection to the mainframe. Everything else on the subnet that the mainframe is on, will connect to the laptop. It is only the single ip address that she cannot ping. When I change HER ip host address (in the same subnet), she can connect again to the mainframe. No one else in eight different routed locations has a problem connecting to the mainframe. When I did a traceroute from her machine to the mainframe, the packet gets to the serial interface of the 2600, but dies at the ethernet interface where the switch is connected to the mainframe.

The only thing I can think of is a hardware address issue with her ip address on the switch in the mainframe location. connectivity after changing the ip address tells me it isn't a route table issue.

Any ideas would be appreciated.

rk

 

[baddos]

The switch won't know the MAC address of the computer on the other network since it's on the other end of the point to point connection(outside of it's broadcast domain).

Being is that the laptop can communicate if you change the IP address, means to me that there is nothing wrong w/ the laptop or it's OS.

Post us the "show ip route" of both routers along w/ the ip of the mainframe and the ip of the troubled ip.

-Bad Dos

 

[pmajors]

Are there any access lists on the interface that the laptop is connected to? it sure sounds like an access list issue.

 

[Randomnoise]

Do a debug ICMP on the Router, or use IP Accounting and then Ping the IBM, your Traffic may be getting there but not getting back.

Find out if any Configuration changes were made before this happened.

Are you using fixed IP Addresses or getting them from a DHCP?

 

[Whoheard]

What is the IP address and subnet mask of the laptop?

Bob

 

[rkmorrow]

I appreciate all of the responses to this thread, and the suggestions.
I am aware that the switch will not know of the hardware address from the laptop across the routed network. I was thinking of the possibilty that she could have gone to the office the mainframe is located. With a change of address to the local network, the (unmanaged)switch could have not let the arp cache die out and may be holding it to another ip address. I have seen this before with unmanaged switches.

I mentioned in the previous post that everyone in the bulding that she is located (except laptop lady), can connect to the mainframe. Everyone including the laptop is on the same subnet and tcp/ip configuration, so it is not a route issue or a subnet mask issue. I change her address to an open address in the same network,(also includes subnet mask!)and she can connect.

There are no access lists on the routers at all, this problem just appreared one day.

All of this points to a hardware address issue in the network the mainframe is on I think. I think the mainframe or the unmanaged switch the mainframe is connected to, has a conflict with her network card mac address, associating it to another address and it has not died out.

 

[adam108]

Are you connecting TN3270 to the m/f?
Unless you are using SNA to connect their is no reason to believe that mac address has anything to do with your connectivity problems. The Mac Layer of the mainframe lan is more or less independant of the mac layer of the laptop lan unless you are running SNA or another bridged protocol. The two routers serve as layer 3 links and prevent MAC problems between systems on the two lans.

Have you checked the routing table of the m/f? Is the m/f running OSPF or RIP or does it use static routes.

 

[Randomnoise]

I have had a similer problem to this a few times. For whatever reason, the Subnet mask that was configured on the host (laptop/pc) was incorrect and therefore, the pings would get as far as the layer 3 device, (Router/Router switch) and no further. The other failure that I have found was that there was no defauly Gateway configured.

The weird thing was that up to that point it had always worked and strangely enough the User had not altered anything (apparently).

 

[crazynoodle]

I know you stated the laptop has the same IP configuration as the other 30 devices on its Lan, but are you certain the Default Gateway for this laptop is correct (1601's Eth interface)? Even if your route information on the 1601 and 2600 is correct - the laptop without the Default Gateway doesn't know what to do with an IP packet from a IP subnet outside its own.. (I appologize if this is a "no sh&%&quot I don't think its an Arp cache issue at the main frame...

 

[Whoheard]

It is really strange that it is just that one IP. Are you sure that you are not violating any octet boundary rules with the ip address? Just checking.

Bob