02-02-2012 06:49 AM
I am comparing two resiliency models for the following scenario. I have 8 local controllers with 7 active and 1 as backup (ie N+1 sparing). All the controllers, including backup, are on a redundant L2/L3 switch. I believe that I have available two resiliency methods.
Method 1: I can create a VRRP instance between each active controller and the backup controller. The backup controller will be running 7 VRRP instances.
Method 2: I can use a Backup LMS IP to redirect to the backup local controller.
Neither method is hitless because I can't put all the User VLANs on the backup controller.
I prefer Method 2 but I would appreciate other views.
02-02-2012 06:53 AM
Personally, in this scenario I would go with the Backup LMS option. Since all the VLANs wont be on the backup controller, you kind of want the clients to get disconnected and reconnect after the failover, so that they get an IP on the new VLAN.
Now, if you do have a VP area, and you can get that VLAN onto the backup controller, I would use a VRRP for that AP group. That way at least your important people stay connected without a drop.
02-02-2012 06:59 AM - edited 02-02-2012 07:01 AM
I do have a VP area and I can indeed get those VLANs on the backup controller. So I will use Backup LMS IP except for the VP Area which will need only one VRRP instance. That is a great idea.