CUCM NIC duplex mismatch

[egua5261]

HI There,

We currently run a CUCM 7.0(2) cluster. I detected an error message on the syslog of the catalyst switch, where the two call managers are connected. The error says this,

"CDP-4-DUPLEX_MISMATCH: duplex mismatch discovered on GigabitEthernet5/17 (not half duplex), with syd-ccm-02 bond0 (half duplex)."

I have checked both the interface Gi5/17 and the nic settings in the call manager's etho and both are set to auto. mode. See below for the details,

Gi5/17 in catalyst switch (4507):

#show int gi5/17

GigabitEthernet5/17 is up, line protocol is up (connected)

MTU 1500 bytes, BW 1000000 Kbit, DLY 10 usec,

reliability 255/255, txload 1/255, rxload 1/255

Encapsulation ARPA, loopback not set

Keepalive set (10 sec)

Full-duplex, 1000Mb/s, link type is auto, media type is 10/100/1000-TX

eth0 NIC: (call manager - MCS7825-h3)

admin:show network eth0 detail

Ethernet 0

DHCP : disabled Status : up

IP Address : IP Mask :

Link Detected: yes Mode : Auto enabled, Full, 1000MB/s

Duplicate IP : no

Queue Length : 1000 MTU : 1500

MAC Address :

RX stats:

bytes : 2365731208 packets : 16624908 errors : 0

dropped : 0 overrun : 0 mcast : 204461

TX stats:

bytes : 4293453743 packets : 35268789 errors : 0

dropped : 0 carrier : 0 colsns : 0

DNS

Primary : 10.0.100.252 Secondary : Not Configured

Options : timeout:5 attempts:2

Domain : 9msn.net

Gateway : 10.0.100.254 on Ethernet bond0

Can anyone please help me to pin point the mismatch issue?

Regards,

Esteban

[#2]

The "CDP-4-DUPLEX_MISMATCH: duplex mismatch " is logged when CDP discovered mismatching duplex configuration on neighboring interfaces.

1. Configure the interfaces so they are both running at the same duplex (full- or half-duplex).

2. Disable CDP

3. Upgrade the ios image to version 12.3 or later.

[Alexander.Khlemanov_2]

Hi Esteban,

We have exactly the same problem with CUCM 7.0(2a) servers connected to 3560G PoE-48 switches. Did you manage to find the solution to the problem?

Regards,

Alex.

[egua5261]

Hi Alex,

Just need to restart the servers in order to sync up with the switch again; it seemed the issue was related to an older version of the IOS in our layer3 switch. Although the duplex/speed were set to auto on both the servers NICs and switchports there was a still a message coming up under the logs of the switch saying there was a duplex mismatch. The other way to fix is by hardcoding these paramaters on both ends.

Regards,

Esteban

[Alexander.Khlemanov_2]

Hi Esteban,

We really prefer to keep them set to auto - the only way how we can use the 1Gb speed. Thanks a lot for the recommendations.

Regards,

Alex.

[xyzjames]

You can only set to Auto on CM server if you want to use 1G. We have this problem on CM702 connect to 6513 on the other side. This is bug. Please see the link

http://tools.cisco.com/Support/BugToolKit/search/getBugDetails.do?method=fetchBugDetails&bugId=CSCsy08872

We have restarted the server several times and the error message still pops on switch.

After disable CDP on the port, the message is gone. But not sure whether CDP is needed for CM7 or not.