C9800-40-K9 HA Pair rebooting

[royce.varughese.1984]

WLC pair of C9800-40-K9 reloads. Noticed errors on RP connection, replaced this and still does same,

Latest we have changed timers to see if it helps

Dec 1 02:27:45.244 WST: %STACKMGR-6-KA_MISSED: Chassis 2 R0/0: stack_mgr: Keepalive missed for 2 times for Chassis 1

This messages keep popping.

Anyone experienced.

[MHM Cisco World]

Are you interconnect RR back to back?

MHM

[royce.varughese.1984]

Yes RP is b2b..

[eglinsky2012]

What software version? Is ROMMON up to date? What type of connection is it, copper or fiber? You said you replaced the RP connections, can you clarify what exactly changed?

I have had a TAC case open for months due to our lab 9800-80 WLCs throwing the same error you provided, but there are no reboots or errors on the RPs. 17.9.3 and 17.9.4. Before and after ROMMON upgrade. We were using a 1’ copper connection back to back between the chassis. Production WLCs that use fiber are fine. They advised a longer cable, I think I tried 3’ but it didn’t help. The case is at a standstill because they want packet captures and debugs during the issue, which I tried to get twice, but they aren’t actually capturing the issue. 

So, I am interested in what your solution ends up being, if it’s not a hardware or cabling issue.

[MHM Cisco World]

Do you insert sfp to one of wlc?

[eglinsky2012]

Following up on my specific issue, we finally got a packet capture, and it shows that the standby is not receiving the keepalives sent by the active (as the error states).

A bug has been filed, CSCwi52278 (per SR 695348639). Again, I'm dealing with a 9800-80 pair and I'm not experiencing reboots, but those are only in the lab, not in production. I thought I'd share anyway in case it is the same issue the OP is experiencing.

@royce.varughese.1984, any updates on your end?

[Leo Laohoo]

Thanks for sharing the Bug ID.  Unfortunately, it is still private.

[eglinsky2012]

@Leo LaohooLooks like it's visible now. The details of being affected on 17.9.3 and 17.9.4 without SMU, and before and after ROMMON upgrade, aren't in there, but I let the TAC engineer know in case she can and wants to add that information.

https://bst.cloudapps.cisco.com/bugsearch/bug/CSCwi52278

[marce1000]

                          Ref : https://bst.cloudapps.cisco.com/bugsearch/bug/CSCwi52278

         - Check if tweaking these settings can be used a a workaround (currently) :  https://www.cisco.com/c/en/us/support/docs/wireless/catalyst-9800-series-wireless-controllers/213915-configure-catalyst-9800-wireless-control.html#toc-hId-307825303

 M.

[Leo Laohoo]

---

@eglinsky2012 wrote:

%STACKMGR-6-KA_MISSED

---

Thanks for that.  I am putting that facility alarm into our NMS and it should send us an email if this log shows up.

[royce.varughese.1984]

I had tweaked the timers and looks like have helped, been stable for past few weeks

[Leo Laohoo]

Reboot both units.

[marce1000]

                                     - What software version are you using ?
  - Log on to HA master and standby and issue the command show version | inc reload
  - Also check the output of dir bootflash:/core/ | i core|system-report

 M.