cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
3736
Views
0
Helpful
5
Replies

9800 HA RP recovery

jmecklenburg
Level 2
Level 2

Hi


I have wlc9800 pair in HA mode.. in RP mode not at RP+RMI mode the chasis 2 have a priority 2 and the chasis 1 have a priority 1, and I want the chasis 2 works as my primary, is in myy primary site...in the documentation the higest number have the priority for my case chasis 2


When the conectividad betwen chasis 1 and 2 is lost, both take the active rol and that it's OK, but when the conectivity as recovery the chasis 2 it's reset with the reason "active unit removed" and the chasis 1 works as active. And it's is not working for me.

 

This behavior it's ok?

fixie rider
5 Replies 5

Arshad Safrulla
VIP Alumni
VIP Alumni

I do not recommend using HA-SSO this way, yours is a perfect use case for N+1 HA where you will have one controller active in one site and another controller active in another site. You will still have AP and Client failover (Not stateful switchover, AP will re-register to new WLC and Client will have authenticate again with new WLC) between WLC's.

More info in the below white paper.

Cisco Catalyst 9800 Wireless Controller N+1 High Availability White Paper

I'm aware to the N+1, just I want to know if the behavior it's OK. Because if both take the active role and the RP connectivity is recovered the chassis that have to reload should be the priority 1 not the 2.

 

Thanks

 

fixie rider

Usually when RP goes down WLC uses active wlc goes to active recovery to prevent dual active scenarios, so it is expected that the active wlc reloads and standby WLC becomes active when the RP is restored. If you need more clarifications please provide below 

What IOS-XE code you are running?

How did u perform the failover test?

Did any of the WLC's go in to recovery mode?

Do you have any console or CLI output from these controllers during this behaviour?

 

Also if you are running a compatible code, what prevents you from configuring RMI+RP with gateway monitoring for HA-SSO?

 

 

The failover test works fine, when are using 17.3.5c, and no when the RP goes down, both controller goes as active as expected.

 

Attach some log from standby when the RP goes down take control as active and then force the switchover.

 

Regards.

fixie rider

As I mentioned previously when RP is lost active WLC will be in active recovery mode (no traffic will pass though this WLC), and standby WLC will immediately transform in to active. So the behavior you are noticing is very much expected

Review Cisco Networking for a $25 gift card