ARP Response Issue

[abdulwajid1]

I am having a strange issue with users connected to Gate Office switch. Users started compaining that they are losing connectivity after some time and getting back after few minutes.
To troubleshoot the issue i put a continous ping to the gateway(192.168.23.1) which is my Core Switch and installed wireshark to capture the packets. When the device loses the connectivity i got "Request Timeout" and "Destinition Host Unreachable". when i take a look at the wireshark packets the PC sends ARP requests on the network to resolve 192.168.23.1 to its mac address but fails to get the resolution. After maybe 10-15 minutes the ARP request gets it resolution and packets start to flow to the gateway. I also checked the arp cache in windows, when the PC is reaching the gateway the ARP entry is there but it gets removed automatically. 
I have tried different PCs but the issue is same. CISCO support says it is an issue with windows but i have tried different new windows machines the issue is the same.
If i add a static ARP entry in windows machine there is no connection problem.
Any idea??

I had the same issue with my new factory switch. Tried alot of workarounds, the workaround that worked was setting the link between New Factory and QC Switch as access not trunk.

[Cisco Freak]

Your PC is pinging the default gateway and all of a sudden it loses the ARP entry from its cache. This doesn't sound good.

What version of Windows are you using? Do you have Linux machine to do the same test.

CF

[Jon Marshall]

When you lose connectivity what do the mac address table and arp cache on the core switch show for the PCs mac address ?

Jon

[Richard Burts]

Are any of the switches (especially the gate office and QC) layer 3 switches that might have routing enabled? Or is all of the routing done on the core?

I have seen symptoms similar to this caused by the difference in arp time out and mac address ageing. Perhaps you could try making the arp time out shorter to match the mac timer.

HTH

Rick