Controllerless Networks

We are about to deploy the Aruba IAP-105s firm wide to the tune of about 200 APs.  I have been having issues with my pilot group where the IAP-105 suddenly fail.  Fail in the sense that they no longer are able to pull Power from the switch.  The switch reports 

 

 

ILPOWER-3-CONTROLLER_PORT_ERR: Controller port error, Interface Gi1/0/15: Power Controller reports power Imax error detected

 

which basically means the switch sees the device trying to pull more power than it can support.  Now I have 10 other APs connected the same type of switch running the same IOS.  The failures have no pattern.  Today I had a 4th one fail in this manner.  Once they fail they will work only if they are powered externally.  

 

I had a second one fail today as well.  This was the one that I used to replace the failed one.  I did plug it back into the same port and it failed about an hour later.  I have had APs fail on other ports so it isn't isolated to a single port or switch. However,  the second one to fail today will work if it is connected to a Cisco 3750-x which supports up to 30mw of power per port.

 

I have changed switches, cables, IOS on the switch and the IAP.

 

Anyone having similar issues or know of anyone that is.

 

 

I am running the latest Aruba IOS 6.2.1.0-3.3.0.1_38408. Everythign else works like a charm,

How many ports that switch has available ?

Do you have any other PoE devices attach to that switch ?

When that happens can you check the total power inline available on the switch ?

The Cisco will give the Aruba AP the maximum amount of power available on that power even if the AP doesn't need it .

This has ocured on both 24 and 48 port 3750s.

 

The switches are not fully populated

 

Module Available Used Remaining
                (Watts) (Watts) (Watts)
------           --------- -------- ---------
1                   370.0 79.9 290.1

 

There is plenty of power unused.  This isnt an issue with the swtich being overloaded.

 

 

You could also try extending initial power delay on the Cisco switch : power inline delay shutdown 20 initial 100

 

 

For certain IEEE Power Devices that require multiple reloads during initialization, you can set a delay shutdown time to configure the switch to continue providing power during initialization.

 

 

Hope this works.

That for the suggestion. Unfortunately my 3750G-24PS does not support that command.  I do have a TAC case open with Cisco.  It seems to be similar to this;

 

https://supportforums.cisco.com/thread/157276

 

 

However the issue is sporadic.  I have 10 or AP that are currently working fine.

I recently had the same issue with an AP that was working fine, along with 15 other AP's, on the same 3750x switch.  I'm running v12.2(55)SE5 code.  I received about 4K messages in a 12 hour period about that particular AP and switchport.  I found this link that suggested making a config change to address the issue. I made that change and so far so good - the AP is up again and the errors stopped.  It's only been a few days so it hasn't baked in fully yet but may be worth a try if your switch supports POE+.

 

https://supportforums.cisco.com/docs/DOC-27050?referring_site=facebook&channel=CiscoSupportCommunity

 

 

Thanks,

 

Ian

We have not yet had that problem with well over 300 iAP installed, but we see it all of the time with the IP Cameras we're also installing.

In our lab we've been able to (mostly) correlate the issue to line-length/termination quality and find that the power inline 2X mode change makes it work in most cases.

Good luck.

I've had it happen to an addtional two other AP's recently (one AP and one IAP) at different locations.  When it happens to us, the AP's have to be replaced to get it to work again.  Changing that setting to 2x doesn't help with our 3750x switches.  Replacing the faulty AP fixes it :)

 

Ian

---

@istong wrote:

I've had it happen to an addtional two other AP's recently (one AP and one IAP) at different locations.  When it happens to us, the AP's have to be replaced to get it to work again.  Changing that setting to 2x doesn't help with our 3750x switches.  Replacing the faulty AP fixes it :)

 

Ian

---

istong, 

 

Please check with support.  There is a specific configuration that might create that issue.

We have a location 12 switches 3750E with version 12.2.(55)SE5 and we have random AP105 (in campus mode) that start rebooting or stop working all together and either replacing the AP or all of a sudden the AP starts to work properly again.

When the AP was rebooting and the IMAX error was reported on the Cisco switch we tried changing the patch cable and changing the port but we were still getting the same issue.

We tried the solution changing that setting to 2x and we still didn’t see any results. We have tried opening a case and were recommended based on old posts from Airhead to disable the WiFi scanning and we still have the issue.

We have replace over 10 AP’s at that location and once we get the AP’s when plug them using stand alone POE we no longer see the issue.  

 

We are still working with SUPPORT on this issue if there is a solution that will work for us we will post it here.

 

 

Good luck!

Check to see if spectrum monitoring is on in any of the radio profiles of that ap-group.  If so, uncheck it:

 

Sorry for the begginer language there but " WiFi scanning" I was refering to that function Spectrum :)

 

FYI and the ticket with Support went to -> Bug ID : 85516 (https://bugzilla.arubanetworks.com/show_bug.cgi?id=85516)

Here is what I think is the problem with the iAP and AP105: http://community.arubanetworks.com/aruba/attachments/aruba/IAP/3936/1/Field%20Bulletin%20AP1311-1%20(AP-100%20Series%20PA).pdf

 

Given that article if the applying the FIX if the AP/iAP is still doing the same thing RMA it and after it should be fine.

We are moving the "FIX" to the production network of our client. I will let you know if we find anything else.

 

 

Good luck

I discovered that this is related, at least for me, to having Background Spectrum Monitoring on,  I had turned it off when working on this.  Things had been going well so I thought the IOS upgrade fixed things.  A few days after I turned on BSM I had 2 APs in different locations fail.  I would check that and disable if it is in.