Every package is running 2.9.40. Not sure what else to include here about the software =)
I have several interfaces, but only two of which are of concern. One interface, office-network, had some PCs that could not ping about a dozen hosts on another interface, customer-network. A routerboard that was located on the customer-network interface, along with the dozen hosts, had no problems pinging them nor did the router with the interfaces mentioned above.
After 10 minutes all but 2 hosts came right back up (note this is why I believe ARP was at fault here, due to the cache timeout). The last two remaining hosts I simply deleted the dynamic ARP entries on the router and both were pingable again from the office-network.
While the hosts being pinged were only APs and simply passing traffic, I am far more concerned if a customer’s IP address would have had the same problem. I am told that the routerboard at the tower site where these devices all sit was rebooted (power issue, routerbord issue, I do not know - yet!) which I believe the problem to be.
My first idea is that the bridge on the four ether interfaces on the tower site wasn’t aware of where the IP/MAC was as it was just rebooted, though that wouldn’t account for the fact it was pingable from the router.
Does anyone have any ideas? Again, I’m concerned that this will affect customers down the road, regardless if a week, month or year away - I’d like to prepare for it!