Hello,
We have a firewall connected to a switch. We have a problem with one of our servers attempting to talk to another server connected to the same switch. For quite some time we had no problems until suddenly about two months ago, they started acting funny. We finally discovered that when one server loses connectivity, it attempts to re-establish commmunications, but upon doing so, it arps the other server and is given the ip address of the gateway instead of the ip address of the other server.
We attempted to fix the problem by clearing the arp and mac address table on the switch and rebooting the server. We verified neither IP has a duplicate entry on the network, and the switchports are seeing the correct mac to IP entries. They said the problem started when we swapped another ASA to which this network connects to an ASAX. That ASAX is designated to the parent network. The arp problem is happening on a separate segment where the servers and switch have their own ASA.
The problem that is occuring is similar to the one described in the link below, only we aren't running LAG or LACP as the connected switch has only a single trunk link to a switch in the parent network. The question, therefore, is has anyone seen anything similar to this and if so, what was the fix?
https://serverfault.com/questions/828426/arp-icmp-issue-in-vlan-over-bonded-interface
