DHCP conflict reported on C880 router connected to MX64 WAN

[AZEIDLER]

Hi ,

We are facing the following situation on a MX-64 deployment

The WAN and Internet connectivity is provided by a MPLS network.

The MPLS router is C819G -4G-G-K9  that was configured to provide the WAN IP address to MX64 (configured as DHCP on Internet interface)  :

!
ip dhcp excluded-address 10.124.113.25
ip dhcp pool MERAKI
network 10.124.113.24 255.255.255.252
default-router 10.124.113.25
dns-server 10.105.160.101 10.105.160.102
!

In this case, the WAN IP provided to Meraki is 10.124.113.26  (the only IP available on this DHCP range)

In a few sites, the Meraki changes the status to "BAD INTERNET"  and the following output is obtained on router, indicating that the IP provided is in the conflict list 

ROUTER >sh ip dhcp conf
IP address        Detection method Detection time        VRF
10.124.113.26   Ping Aug              09 2018 11:21 AM

The MX64 WAN IP address respond to ping for a period of time, and then stop responding.  At this time, the Dasboard reports "UNREACHABLE"

The situation is resolved whent the DHCP process is restarted on routrer.

Any Idea what can cause the DHCP conflict ?

[PhilipDAth]

The two DNS servers definitely respond ok?

Any chance something else has been plugged into the back of the 891 and is grabbing those DHCP addresses?

[PhilipDAth]

When it is broken, what is the output of these commands:

show ip dhcp bindings

show ip dhcp conflicts

[AZEIDLER]

Hi.

Below the outputs:

ROUTER r>sh ip dhcp conf
IP address       Detection method Detection time               VRF
10.124.113.26  Ping                     Aug 09 2018 11:21 AM

ROUTER>sh ip dhcp bind
Bindings from all pools not associated with VRF:
IP address Client-ID/                         Lease expiration               Type
                   Hardware address/
                   User name

[AZEIDLER]

Hi,

No chance of another device pluggd. We have built a specific VLAN for WAN connection with only the Router and the MX64 connected to it;

Not tested DNS during failure, but since it´s a very large network, DNS servers should be reliabel 

[PhilipDAth]

Are you sure the WAN interface on the MX has now been configured with a static IP instead of DHCP?

[PhilipDAth]

Assuming the WAN interface is definitely using DHCP and is not configured with a static IP address I think my next steep would be to configure a larger DHCP pool and clear the DHCP conflict database.

I would want to make sure that nothing else allocated a second IP address from that pool.

[PhilipDAth]

I also assume al these 891's are running the same IOS version?

[AZEIDLER]

After this issue happened on several devices we have changed the WAN to static (on same IP) and the problem has not happened again.