Mobility ping and SSH errors after upgrade to 7.2.110.0

[mduling]

After upgrading my 5508s to 7.2.110.0, they are reporting mobility data path errors to one of my WiSMs running 7.0.235.0.

I get these messages on the 5508s reporting that it can't send a ping to the affected WiSM:

*ethoipSocketTask: Aug 08 21:15:41.175: %ETHOIP-3-PKT_RECV_ERROR: ethoip.c:341 ethoipSocketTask: ethoipRecvPkt returned error

*ethoipSocketTask: Aug 08 21:15:41.175: %ETHOIP-3-PING_RESPONSE_TX_FAILED: ethoip_ping.c:312 Failed to tx a ping response to <ip address>, rc=5

But maybe there is another clue because I also see in the same log these errors referencing the same WiSM:

*bcastReceiveTask: Aug 08 21:15:45.310: %LOG-1-Q_IND: mm_dir.c:1969 Failed to recreate the SSH Rule for <ip address>.

*mmSSHPeerRegister: Aug 08 21:15:44.829: %MM-1-SSHRULE_CREATE_FAILED: mm_dir.c:1969 Failed to recreate the SSH Rule for <ip address>.

Why is the controller trying to SSH to another controller?  Was some SSH related feature added to 7.2 that has been accidentally enabled?  Any insight into this issue would be appreciated.

[Amjad Abdullah]

You need to ave both controllers in same version if they are

On same mobility group. Those errors are expected if you aren different versions because the mobility link can not be correctly established.

HTH

Amjad

Sent from Cisco Technical Support iPad App

[mduling]

That isn't my understanding.  My understanding is that compatibility works within major version numbers.

http://www.cisco.com/en/US/docs/wireless/controller/5500/tech_notes/Wireless_Software_Compatibility_Matrix.html#wp89677

[dkleimbach]

I'm getting this same error between 2 of my 5508 controllers are they are both running the same version.

*mmSSHPeerRegister: Aug 08 21:15:44.829: %MM-1-SSHRULE_CREATE_FAILED: mm_dir.c:1969 Failed to recreate the SSH Rule for .

Did you ever have any luck on solving this one?

[mduling]

My memory may be incorrect, but I think the SSH rule error was a cosmetic bug and not related to the main problem, which I do remember well.  Namely, the mobility connectivity problems.  Those problems began after shifting a number of APs onto a lightly loaded controller.  The mobility issues baffled everyone, and I finally just moved the APs back off the affected controller and the mobility issues went away.  Moved them back again and the mobility problems did not return.  It must have been some bizarre fluke.