ARP issue on Cisco Switch

[fgasimzade]

Hello everyone

We are having some strange arp issues in one vlan. Suddenly, some devices (not all of them) from vlan 1 are not able to ping vlan 2. Interface vlans are located on different switches and I have to manually clear arp in vlan 1 to make computers ping each other between vlans.

We have set arp timeout command in vlan 1 but it does not help until I manually clear it

What can be the reason?

Thank you

[nspasov]

I have seen this before and both times it was a bridging loop caused by spanning-tree misconfiguration. Do you have a network diagram on how the network is connected? Also, can you share your configs?

Thank you for rating helpful posts!

[fgasimzade]

Dear Neno,

We have mst in place, but vlan 1 in our case is located on a root bridge and switch with vlan 2 is directly connected to it. I also suspected STP but this is happening with a root bridge and only for these 2 vlans with some computers. All other vlans are accessible

[nspasov]

Hmm, have you:

1. Configured Root guard on all of your root bridge's ports?

2. Confirmed that ports set at portfast have pbduguard enabled

3. Use show and debug commands to make sure that TCNs are not happening for the affected VLANs

4. What model switches do you have and what is the version of code are they are running

Thank you for rating helpful posts!

[fgasimzade]

Update:

Affected computers (for now these are just 2 mail servers located in vlan 1) are not able to ping any vlan on switch 2. Nothing is blocked by STP.

Mail server 1 is connected to Root bridge vlan 1, mail server is connected to another switch in vlan 1. Both servers are configured as a Cluster with Microsoft Exchange 2010. I suspect this could be a cluster issue, but not sure

[fgasimzade]

One more update:

Whenever this happens to mail servers in vlan1 I can ping vlan 2 from my computer and my computer is also located in vlan 1