Joe, I had the exact same problem but in a larger environment. Some client on one subnet would get disconnected while other did not. After hours and days of troubleshooting and this is what we found out.
1.) As400 needed some PTF files applied for the TCP/IP stack (Basically applied patches to the AS400) 2.) Also the NIC on the 400 was running 2 frame types 802.2 and 802.3. What we found in the sniffer trace was the AS400 would send out a 802.3 frame every once and a while and cause the switch to re-arp and client would lose the telnet session to the 400. I cannot pin point a single client or segment I did not know how it dropped this person and not this one. Once we remove one of the frame type to stop the re-arp. No more problem. No I cannot say the PTF applied to the 400 did not fix the problem also. Some thing to look at. I am curious to see if anyone else has had a problem? Ronnie ""Joe Morabito"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > Hi all. I have a real world problem and would like some thoughts, here it > goes. > > Client has a central site with 7 remote locations. Central site has a 6513 > with MSFC2, all sites have T1 in. Eigrp is used between sites and is fine. > At the central site lies an AS/400 that all remote sites connect into (users > use IBM client access). All is IP no SNA anywhere. What happens is at one > and only one location users get bounced out of there session about 4 or 5 > times a day. They have full connectivity to the central site, just no > AS/400. > > So I was at the remote site where the problem exists, could ping the as400 > fine. So I waited for a "blackout", it came. When the users were bounced I > tried to ping the as400 and was unsuccessful. I tried to ping the central > router and was good, I tried to ping the vlan ip address that the as400 was > in, that was good. I even tried to ping other computers in the same subnet > as > the as400 and that was good. Tried to ping the as400 again and that was NO > good. These outages only last a couple of minutes and then it comes back up. > > There is NO route flapping. All eigrp neighbors stay put and all routes stay > put. > > What is really strange is that this happens only from one remote location, > all > other locations are fine. The as400 has only one NIC (ehternet) and all > other > sites connect to the same IP address. > > So where is the problem? I checked the port the as400 was plugged into and > found some errors, I found out that the port was set to 100/full and the > as400 > nic was set to 100/half, so I adjusted the port on the switch to 100/half. > But if this was the problem, wouldn't all sites have trouble? Could the > switch possibly be aging out its cam table to the remote site that is having > problems? > > I don't remember if they blacked out again after changing the duplex (I had > been staring at debugs for about 8 hours). > > I called TAC and bumped up the outbound queue size to the site that is having > problems, but haven't had a chance to call them back. > > Any suggestions? > > Thanks. > > -Joe > > [EMAIL PROTECTED] Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=42399&t=42391 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
