4431 vrf Mgmt-intf arp cache?

[curtmcgirt]

this was a running router without g0 configured. i enabled it this way

interface GigabitEthernet0
description MGMT_INT
vrf forwarding Mgmt-intf
ip address 10.2.0.5.255.255.255.0
ip access-group MGMT in
negotiation auto
end

ip route vrf Mgmt-intf 0.0.0.0 0.0.0.0 10.2.0.1
ntp server vrf Mgmt-intf 10.0.0.10

when i first configure this, i can (ping vrf Mgmt-intf ) both the next hop and the ntp server. if i immediately try again, it works, and keeps working until i stop trying and let it sit there for 30-ish seconds. after that, subsequent ping attempts fail, until i "clear arp". then ping works again. temporarily. it works long enough to get the ntp server as a peer. temporarily. (the behavior also persists if i remove the access-list from the interface)

there was formerly a static route to 10.0.0.0 through another interface, but i removed it. 

any thoughts on what might be causing this? i don't see anything suspicious in show arp.

edit: also the only IPs in show ip arp vrf Mgmt-intf are 10.2.0.5 and 10.2.0..1

[marce1000]

 - Have a look at the current  software version being used, make sure it is not too old and or compare the problem when an advisory release is  being used.

 M.

[Georg Pauwen]

Hello,

odd. What is the uptime of the router ? If it is very long (> x years) a simple reboot might help...

[curtmcgirt]

The router was rebooted on October 31 during an upgrade to Cisco's currently recommended OS version. 

[curtmcgirt]

sigh. documentation indicated i was working with a subnet untouched by human hands. that was not the case. 10.2.0.5 was a duplicate ip address.