11-22-2012 11:24 AM
I'm not entirely clear on what the 'tracking master-up-time' command actually does inside a vrrp declaration. I've tried reading the User Guide and Command Reference, but I'm still not sure. It appears that it ups the priority of the current VRRP master by the specified 'add' value after it has been the master for the specified number of minutes.
For example (showing only the timer bits):
tracking master-up-time 30 add 20
... this means that this VRRP instance has a normal priority of 100. If (and only if) this controller becomes the master and stays that way for 30 minutes, the priority will rise to 120.
As such, this function appears to be useful to make a current VRRP master 'sticky' and avoid unnecessary transitions to another, once it has been the master for a period of time.
if we just want IP (and the APs using it) failover from one controller to another and then preempt back, when it returns, there is no need to use this option? It's only useful when you have multiple controllers in a failover group (using VRRP) and want to carefully assign different APs to each controller, during failover?
Am I right here?
Solved! Go to Solution.
11-22-2012 12:10 PM
It is designed to ensure that a master will only be allowed to take and maintain control of the VRRP if it has been up for a certain amount of time. There may be a time that the controller that you want to have control is flapping (going up and down) for whatever reason. You only want a device to become and stay the master control only if it has been up for a certain amount of time.
Planning what controller your APs attach to and fail back is separate from this.
Quite frankly, I have never used this option. I would only introduce it if I think I have some sort of random hardware failure and I don't want a partially failed device to take control of the VRRP.
Aruba Customer Engineering
Looking for an Answer? Search the Community Knowledge Base Here: Community Knowledge Base
11-22-2012 01:05 PM
OK - thanks. I've just done some more tests - to confirm:
The up-time is measured as the time since the master became the VRRP master. After this time, the controller will raise its priority by the specified amount.
It's not since the VRRP instance was 'no shutdown'ed (and a backup) or the controller itself came up. In other words, it only has an effect when this controller is the VRRP master.
The reason I ask is the command was put in there by our integrator, when the system was first set up several years ago - all I really want to happen is things to fail over to a backup, when the master fails, and return to the master when it comes back. We're just setting up a couple more locals and I think we can safely omit it from the new VRRP groups and remove it from the existing ones.
Thanks again (but do reply if I'm wrong!).
11-22-2012 01:09 PM