Bad Arp Entry ... I think

Having a weird issue that I can't seem to sort out, wondering if anyone might be able to send me in another direction.

I have 2 switches out of 40 that cannot be reached by Cisco Prime.

Switches are on VLAN42, Prime is on VLAN30, L2 Network routed through N7K cores.

When I check the ARP tables on the switches that won't communicate, they have entries for the Prime Server, on VLAN30.

They have the correct IP Address, correct MAC, and the VLAN that the Prime Server resides on.

​

When I clear the ARP cache, the switches can communicate for a few days, but eventually the bad ARP entry comes back and they lose connectivity again.

I've looked at the ARP tables leading back to the cores and nothing seems fishy, I'm not totally sure what I could be missing here.

Not looking for someone to solve my problem, just maybe light a fire in my troubleshooting.

Something seems off that there is an ARP entry for a device on a different VLAN, but I don't know if that is normal, I can't find much information on how ARP is handled when VLAN's are involved.

TIA!