Controllerless Networks

We have a school district where one location, a small charter school, that only has 4 Instant APs is experiencing constant reboots.

 

When we review the logs we see several puzzling errors:

 

• Cannot find AP 192.168.18.246 (it's own IP address)
• Reboot reason: Master transitioning to local

 

They are all on a flat network with no VLANs.

 

This has been happening since we installed them and Ive been working with TAC for aout 2 weeks with no reolsution.

 

Anybody experienceing this?

 

 

The ususal questions:

What model of AP?

What version of code?

 

Several version ago, we had a similar symptom - 6.1.2.3-2.0.0.2 or maybe one earlier - the iAP ARP protocol and our network switch vendor (Adtran) ARP protocol interpretations were subtly different.  The result was that each AP would see the others, then as ARP table entries aged, they would loose "sight " of one another.

 

We had to shut them all down, then bring one up, upgrade the code and power it back off and do the next. Once they were all on the next rev (an engineering build candidate released early for our problem) we brought them up and the ARP issue was fixed.

 

 

AP-104s and one AP-105

 

3.2 - latest build from support site (6.2.0.0-3.2.0.1_36986)

 

All connected to an S2500

Yep i vote for that upgrade it... you will have to do it via console via TFPT...

 

Here is the process

 

We're on the latest rev.

 

6.2.0.0-3.2.0.1   36986?

Yep. See below.

 

yeah sorry i just saw it sorry for not reading...

 

go to the console of the instant AP

 

do

clear os

purgeenv

saveenv

reset

That will bring it to the firmware that it came from the factory... 

 

Try this with one of hte IAP of course turn the others off....

if it works do the same with the others.. and then start including them to the cluster one by one..

 

You will loose your config though... but well instants are really fast to configure..  might be a firmware error i mean that it didnt upgrade correctly

I have even placed them in there own Management VLAN and tagged the clients for the school VLAN and they still reboot.

you did clear os?

If they still do then i would tell them for a RMA...

okay try that!

IF it doesnt work... then tell the TAC That you would like to do a RMA and tell them all you did and nothing... because well its affeting your network and you cant continue with this...

Yes, that was my next plan of action. 

 

Wish me luck...

• Any chance there is an IP conflict with that 192.168.18.246 address?  Have you tried changing the virtual controller IP to something else?
• Do you have the ablity to take the S2500 out of the mix?    Can you put the IAPs on their own small switch; even if it is just off hours to ensure the S2500 is not to blame in some way.

True that Clembo.... that could be the issue....

 

We placed them in their own management VLAN with DHCP from the S2500 so no IP conflicts. Changed the VC address to the VLAN subnet.

I've factory reset the APs today as well.

Last step is to bring in a different switch and see if there is any change.

haha you did all at once...

We will not know what was the issue... but oh well i hope it does not bother you anymore...

 

Your case its weird because i have deployed on that exact firmware on some clients with no issues...

 

I told you to factory reset so you could start with the factory default firmware because i once had an issue in which i was not able to connect to the AP, the managment was going on and off... just many weird things at the same time... i though it was something related to the firmware as it didnt make any sense... though it could be a bug or a bad firmware upgrade... Anywyas i fixed it that way, at least that time.

I am seeing the same issues in a 4 IAP deployment. It might be firmware related.

ArubaInstant_Orion_6.2.0.0-3.2.0.2_37229 seems to have fixed it :)

TAC just sent us that today. We've upgraded.

Let you know tomorrow if it fixed it for us.

Hi guys, I have the same issue with 13 IAPs, with the last version. They seems to reboot, and they give me the same error message or reason: 

   AP uptime is 1 day 12 hours 22 minutes 47 seconds
   Reboot Time and Cause: master transitioned to local (from 6c:f3:7f:cb:db:6b 10.10.0.21)
   Core file URL: http://10.10.0.25/core

 

I'm thinking about power off the master, just to see if it is a VC issue. 

The IAPs have been working one week, and the only issue is that they reboot.

AgentED, did you solve the issue.?

Nope. :-(

We installed it two days a go and we've been rebooting as normal.

No change.

Just checked our AirWave server and two of the APs have been up for 6 days while two rebooted 8 hrs ago.

Arg, I was hoping the update would fix it.

Well, at least 50% seem to be ok now I guess there's that.

I'm goint to open a Case with Aruba, just to know if it is a known issue that could be solved with a new SO version. :smileysad:

Hi guys.!! 

 

  Instant',s next release is going to be available soon, so, let's wait for this update to know if it is the issue reason. 

  An Aruba's engineer is sending me the file to try to solve the issue. I'll tell you how it finishes. :smileywink:

 

Regards,

 

Katherine Villalobos

Just so you guys are aware.....

 

We were experiencing the same issues with multiple of our clients IAP clusters rebooting for no reason. Our only solution to fix the problem was to roll back to the IAP firmware which was released pre airgroup features. 6.1.3.4-3.1.0.1_35899

 

Did this last night. Just logged in and all the APs have been up for over 24 hours. 

 

People, this is a record!

 

We received a new build today that is supposed to fix this issue. I think I'll wait a day, or so before trying it.

 

For right now, they are functioning with no issues.

Please let me know how it goes with the new build. We received one from TAC (C_ArubaInstant_Orion_6.2.0.0-3.2.0.2_37649).

 

Not sure if I want to be the first to try this one. If you upgrade one of your clusters let me know how it goes. 

I just received the new build.!! Who will be the brave to update the IAPs first.? :smileyvery-happy: 

 

New c-build does not work. Rolled back to 6.1. It's pretty clear now that the issue only exists in 6.2.

We're staying on 6.1 until this is actually resolved.

We are currently rolling all sites back to 6.1.3.4-3.1.0.2_36584

Hi guys.!!

 

   I rolled back this deployment to version 6.1.3.4-3.1.0.2_36584. But after that, I got a new issue.

   I have two guest networks using Captive Portal with email. when not authenticated users go to any http page, they are redirected to Captive Portal. But, if they try to go to a https page, they get a certificate error message and then, theyare redirected to the Virtual Contoller login page, instead of the Captive Portal. 

 

Do you have experiencied something similar.??

 

Regards,

 

Katherine 

Kavilla07,

 

The further back you go, the more bugs you will probably introduce.  Did you open up a case for your original issue that it can be fixed?  We want to make sure that your problem is fixed before referring you to code that will expose you to more issues...

Hi.!!

 

  Yes, I worked on this issue with a case. The recommendation was to try the new release, 6.2 but it cotinues with the problem. So, the new advice was to roll back to version 6.1.3.

 

 

Regards,

 

Katherine V.

Does anyone know where to get the 6.1.3 firmware? I seem to be unable to find it on the Arubanetworks site.

It's on Aruba's support site. You need to go under "Aruba Instant" and then "Current Releases". 

 

I see 6.1.3.4-3.1.0.2_36584 as a downloadable option. 

Wow, thanks for your quick reply.

 

We do, however, have the Dell labeled IAP 93s.

 

I assume that won't be a problem when I overwrite the current Dell branded firmware?

 

Again, thanks for your reply.

I've seen several warnings here abouts that the Dell gear needs Dell firmware.

Someone correct me if I'm wrong, but I'd think twice before attempting to re-brand your equipment.

Dell does need Dell Firmware.  Please contact Dell to find out if they have something, but have not posted it yet.

 

I just contacted Dell, they will send me a new device.

 

Thanks for the info guys.

Where you able to see which was the issue?

You r last message is in blank.

Just fixed a customer with this problem by upgrading to 6.2.0.0-3.2.0.4. check the release notes for 3.2.0.4 under "Fixed Issues in this Release" -> Access Point Platform.  To identify the issue addresses by this fix, ssh to an IAP and execute the comand show version. you will notice the reboot reason is master transitioned to local.

 

Dining_Hall_East# show version

Aruba Operating System Software.

ArubaOS (MODEL: 105), Version 6.2.0.0-3.2.0.2

Website: http://www.arubanetworks.com

Copyright (c) 2002-2013, Aruba Networks, Inc.

Compiled on 2013-02-15 at 22:20:03 PST (build 37229) by p4build

 

AP uptime is 26 minutes 58 seconds

Reboot Time and Cause: master transitioned to local (from 6c:f3:7f:cb:dd:47 10.10.24.147)

3.2.0.4 fixed the issue but introduced another, after being stable for 4 hours the the aps began to reboot again with reason "image sync fail". upgraded to 3.3 and aps have been stable for 12 hours.

Our network of 24 IAP-105s has been plagued by the "master transitioned to local" reboot problem for a long time.  Glad to see it's finally been addressed!  I'll be rolling out the 6.2.1.0-3.3.0.0 firmware tonight.

3.3 has fixed this issue for us.

Thanks for the update Ed. I have 3.3 in my lab but wasn't sure it was safe to start rolling out on production sites. 

3.2.0.4 fixed this for my organization. We only had one AP rebooting at random and has not done so since the upgrade.

Where do I get 3.3. ?

v.3.3 is available for download on the TAC Support site under "Software Downloads".

http://support.arubanetworks.com login or create one and go to "Download Software"