Wireless Access

Reply

Cluster - New Standby Controller for APs/Clients

I have a cluster of 4 7240XMs in our new ArubaOS 8 Environment. I know that f the current Active Controller becomes unavailable for APs/Clients, they will fail-over to the Standby-Controller - but at what point does a new "Standby Controller" get created/assigned to the APs/Clients that failed over - or does it?

Re: Cluster - New Standby Controller for APs/Clients

The cluster leader will elect the new S-AAC based on load / availability



Thank you

Victor Fabian

Pardon typos sent from Mobile
Thank you

Victor Fabian
Lead Mobility Architect @WEI
AMFX | ACMX | ACDX | ACCX | CWAP | CWDP | CWNA

Re: Cluster - New Standby Controller for APs/Clients

Thanks Victor. One more question. We rolled our ArubaOS 8 Upgrade into our "7240 to 7240XM" controller upgrade/migration project. Started out with ArubaOS 8 straight on 7240XM and did a three week migration plan. We ended up in a situation where we were limited on "10G Uplinks" - so gradually we've been moving them over each day.

 

Remove 1G uplink from Active Controller - Clients/APs fail over to Standby Controller seamlessly - Install 10G Uplink. Was curious if thoughts on a better way of upgrading uplinks on an already in-production clustered controller that already has clients/aps terminated on it - like example if removing the active controller from the cluster profile - would that cause clients/aps to auto-move to the standby controller? Or if what I've been doing has probably been fine? Once I get the last two controllers up to 10G - I'll be adding our redudnant links-port-channels back into the mix. One step at a time.

Re: Cluster - New Standby Controller for APs/Clients

Yes APs and clients should dynamically move to the nodes but I suggest to do it during non peak hours



Thank you

Victor Fabian

Pardon typos sent from Mobile
Thank you

Victor Fabian
Lead Mobility Architect @WEI
AMFX | ACMX | ACDX | ACCX | CWAP | CWDP | CWNA

Re: Cluster - New Standby Controller for APs/Clients


@Victor Fabian wrote:
Yes APs and clients should dynamically move to the nodes but I suggest to do it during non peak hours



Thank you

Victor Fabian

Pardon typos sent from Mobile

Thanks Victor - and so true on the "during non-peak hours". I've been doing it about 7:00AM each morning (giving cluster a day to adapt). Just have one more left after today. The stand-by fail-over as worked pretty well with client sessions (only one dropped ping) for one client I test with each morning - but I'm not that crazy to do it during the day. :-) 

Search Airheads
cancel
Showing results for 
Search instead for 
Did you mean: