Controllerless Networks

We have had multiple instances that clusters do not work, usually either after an upgrade, complete reboot, or when trying to use the Airwave GUI config.  When we SSH to the AP's we get the AP is in a degraded state.  In order for us to better troubleshoot this I was wondering why causes the entire cluster to go into a state like this?  The last time we had to unplug AP's until we found the lucky AP and then the cluster came up.  After it came up then we could power up the other ones and things came up.  My question is there has to be a better way to troubleshoot this state as now the power to the building was lost and now the AP's all came up default (thats conveinent) and the minute we try pushing the Airwave config to it the cluster goes into degraded state.  So we are going to go back out there and unplug all the AP's until it comes back up but I feel there has to be a better way.  Any suggestions?

- What version of Instant OS are you running?

- What access points are you using at that site?

- What version of Airwave?

We just updated to the latest to hopefully help (It hasn't).  The Client also wanted to utilize the timed SSID feature.

6.4.4.4-4.2.3- Instant

AP-205

8.2.0- Airwave

Just to understand:

The cluster was in that state; You upgraded and it is still in that state;  Did it ever recover or it never recovered?

No problem.  Earlier this week we updated the code and when it came back up from the upgrade it came up in degraded state.  We got it to come back up (on the older version) by unplugging the AP's until the cluster came up.  Once it came up we then plugged all the AP's back in.

The cluster came up and operated but was on old code.  We then upgraded it and it all came up and worked great.

Of course that same building had a power outage for a day.  When the cluster came back up the cluster was defaulted (Makes no sense) we then tried to push the config to it from Airwave.  Once we did it went into degraded state again.  We have now got it to come back up but it is defaulted and we aren't really sure the next best steps.  My guess is one AP is making everything mad but we have no way of figuring out which one.  Does this help?

It sort of does.  You probably need to open a TAC case so that they can run through all the steps with you to understand what might have gone wrong.  If all of the APs were accounted for, and there was no stray AP that might have come up first and defaulted all of the other APs, we have something that should be looked into.  One thing you can do is to set a "Preferred Master" so that even if a single AP comes up first defaulted and takes control, the preferred master AP would take over with its own provisioned config and restore it to the cluster:  http://www.arubanetworks.com/techdocs/Instant_41_WebHelp/InstantWebHelp.htm#UG_files/CustomizeIAPParams/Master_Election_Protocol.htm?Highlight=preferred