Problems with V2110, synchronization and backup tunnel

  • 0
  • 1
  • Problem
  • Updated 2 years ago
  • Solved
  • (Edited)
Hi,
we have two V2110 controllers in 9.21.08 in HA to manage our 800 extreme ap's with 400 local aps in each controller.
we have monitorig the controller with netskght in order to receive all the traps from de aps and controllers. 
we have receiving alerts with lost backup tunel  from the master controller to its foreing aps, but in the slave controller all is normal and all the aps are working fine.
Anyone have this deployment?
we have 30 sec to tunel timeout. the 
The doen message and up message with 1 sec delay.
Device: 192.168.29.21
Severity: Critical
Message: STRING: "14465086085F0000" STRING:
"AP name  'V-BIB-PRIMERA_1A'::
disconnected from the system(poll timeout)"
Trap: apTunnelAlarm
Device: 192.168.29.21
Severity: Critical
Message: STRING: "14465086085F0000" STRING:
"AP with Serial Number  '14465086085F0000'  has been connected to controller"
Trap: apTunnelAlarm
Photo of FES

FES

  • 1,360 Points 1k badge 2x thumb

Posted 2 years ago

  • 0
  • 1
Photo of Umut Aydin

Umut Aydin, Escalation Support Engineer

  • 2,290 Points 2k badge 2x thumb
Hi,

please take a look in to the logs of the Controller .
What do you get here?Which message.



If not please set the AP and AC log level to informational and configure a Syslog Server in order get more info's in the Controller logs. (Controller-> Logs)

Please set one of the Ap to factory default over the AP CLI  ( cset f ) and delete the AP from the WebGUI as well.( AP Bulk conf)

Let's see if this maybe helps.

Check if you enabled secure connection - Controller->Network->Secure Connection.

Other option would be if you have 2 DHCP Server with overlapping IP scope for both COntroller?!

I assume HA is up.

If nothing helps please open a case and GTAC should work on this.

Regards
Umut
Photo of Chat

Chat

  • 298 Points 250 badge 2x thumb
Ansewrs down:

please take a look in to the logs of the Controller .
What do you get here?Which message.

>>>>>On the controller logs we can't see nothing about this fail.

If not please set the AP and AC log level to informational and configure a Syslog Server in order get more info's in the Controller logs. (Controller-> Logs)

>>>>>Wireless Controller Log Level:  Information   
>>>>>Wireless AP Log Level: Information 

Please set one of the Ap to factory default over the AP CLI  ( cset f ) and delete the AP from the WebGUI as well.( AP Bulk conf)

>>>>> This is another problem, a lot of times, some APs is down on controller  but reply to ping                   and              you can do a "cset factory" and the IP come to controller up again, but you                   can't see logs of            fails.

Let's see if this maybe helps.

Check if you enabled secure connection - Controller->Network->Secure Connection.

>>>>> Yes we have Secure conection between controllers.

Other option would be if you have 2 DHCP Server with overlapping IP scope for both COntroller?!

>>>>> Only have one DHCP server per campus. Not overlapping.

I assume HA is up.

>>>>> HA is up. 

If nothing helps please open a case and GTAC should work on this.

>>>>> Thanks, we'll open a case on GTAC.
Photo of Chat

Chat

  • 298 Points 250 badge 2x thumb
Is this normal only this logs when you download AP-logs via ftp server?? this is not very helpfull

"Creation Date: Fri May 13 10:51:32 2016
Software Version: V2110 Large-09.21.08.0013"
Photo of FES

FES

  • 1,360 Points 1k badge 2x thumb
Hi, we still have traps messages from de both controllers in HA with lost backup tunnel from the aps. Its randomly. We can see that sometimes the aps change de tunnel from its primary controller from the backup controller...
Somebody have this behaviour with V2110 and HA? Somebody have upgraded to 9.21.09 ??
Regards
Photo of Raffi

Raffi, Employee

  • 1,598 Points 1k badge 2x thumb

I would recommend to open a gtac ticket to track this issue.

A  tech support log from each controller, and network topology diagram showing how aps connect to controller would be nice.

Then can take a look at some ap trace files from a few aps that recently had a poll timeout.