02-21-2014 02:14 AM - edited 03-07-2019 06:20 PM
hello community
i have a strange behavior in my switch environment..
i have a 3750x switch stack which is the core switch in my network
there are some 2960s client switches with port-channel connected to that core switch
the core switch have different vlan interfaces. vlan1 for workstations and servers, vlan506 for management.
the client switches only have an management vlan interface (id506) - the native vlan1 is shutdown.
as i say, servers and workstations are located in vlan1. when i connect from a workstation, which have an ip address from vlan1, to the client switch and take an configuration backup using tftp on that workstation, the client switch insert an arp entry for this workstation - located in vlan1 - where the client switch dosent have an ip address..
when i troubleshoot this problem, i first see the mac address from the core switch vlan 506 interface and the ip address from the client in the arp table.
after few minutes the switch change the mac address to the real mac of the client..
this is strange because of arp.. the switch should not have arp entries from a layer3 interface in which he dosent have an ip adress.. am i right??
thank you in advance for your help, best regards and stay happy!
michael
03-07-2014 05:27 PM
Micheal,
I ran in to some wird issues like this last year.
Can you try something for us.
On the client 2960 can you change the type of default gateway.
!
no ip default-gateway 172.25.6.254
!
ip route 0.0.0.0 0.0.0.0 172.25.6.254
!
Its worth a try
03-09-2014 11:48 PM
hi Acampbell
thank you for the tip.
i dont have activate the sdm lanbase-routing template on the client switches.
its still possible to work with default gateway.
best regards - michael
03-30-2014 02:38 PM
I wonder if you have made any progress on this issue?
HTH
Rick
03-31-2014 12:38 AM
hi Rick
Sorry..
i have opened an tac case and they are able to reprocedure the issue..
as i have a solution, i update the threat!
best regards - michael
03-31-2014 06:52 AM
Michael
Thanks for the update. I have been thinking about this and wondering if there is a difference between how router IOS treats ARP and how the switch Catalyst IOS treats ARP. With router IOS (and I believe in general how we expect devices to treat ARP) if the router receives an ARP request in which the source address is not in the subnet of the interface which received it then the router will reject the ARP request and not process it. I wonder if the switch IOS is operating differently. When you begin the backup the switch is communicating through the core switch. And I wonder if at some point the workstation sends an ARP response with its IP and MAC and whether the switch does process that ARP response and add it to its ARP table, even though its IP is not in the subnet of the management interface?
HTH
Rick
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide