I have a few questions regarding the new high availability feature in 6.3 and above. I understand it fairly well, but have not configured yet, but looking to do so for some customers after upgrades. Just wanted to be clear on a few things, and hopefully someone else may find the answers useful.
- ha groups are configured on the master and pushed down to the locals?
- ha group membership must be configured on each respective local?
- a controller can only be a member of one group?
- understand the ha backup cannot be a master-vrrp. Realise this is addressed in 6.4, but are there plans to fix in a 6.3 release? That said, would it work if the backup-lms was the master-ip instead of the master-vrrp?
- are there any specific troubleshooting commands to check the status of ha, such as checking the aps connection to its backup is fine etc?
- bridged ssids not supported for ha in 6.3, but any issues with that controller in a ha-group? Is it just the case that tunnelled aps do fast failover but the bridged ones do the traditional rebootstrap?
So in an N+1 deployment with backup-lms as follows. Arrows show direction of failover.
The ha-group would look like this?
ha group-profile ha-group
controller <A-ip> role standby
controller <B-ip> role active
controller <C-ip> role active
controller <D-ip> role active
I have another slightly different setup for a customer that looks like this. Arrows show direction of failover.
The ha-group would look like this?
ha group-profile ha-group
controller <A-ip> role standby
controller <B-ip> role active
controller <C-ip> role dual
controller <D-ip> role active
Thanks