Wireless Access

I have to reboot some of my APs after some network issues since they no longer appear in

show ap active

but they still appear in    show ap database long     with Status  down *

Unfortunate, just error msg to

apboot  ip-addr ...

AP is down 

But the same APs with static IP addr are reachable

ping  ....  ip-addr

!!!!!
Success rate is 100 percent

Any hints how to recover from this situation ?

* points to the backup controller IP addr, not the one of aruba-master

Also AP power-cycle does not seem to help

Running ArubaOS  5.0.4.13

The access points needs to be up for apboot to work.  You should try to drop the power to that access point using the POE switch.

Hello,

AP is down and not power at all or is rebooting and trying to pick up a license? 

Do you have enough licenses in both controlles? 

Regards,

Giovanni Saldarriaga

I noticed some of the same behaviour in earlier releases when a RAP would connect to the controller and when you do the "show ap database" under flags I would see the "D" - letter.

Flags: U = Unprovisioned; N = Duplicate name; G = No such group; L = Unlicensed

I = Inactive; D = Dirty or no config; E = Regulatory Domain Mismatch
X = Maintenance Mode; P = PPPoE AP; B = Built-in AP
R = Remote AP; R- = Remote AP requires Auth; C = Cellular RAP;
c = CERT-based RAP; 1 = 802.1x authenticated AP; 2 = Using IKE version 2
u = Custom-Cert RAP
M = Mesh node; Y = Mesh Recovery

That we would be able to FIX :) only with a power cycle on site.

Interesting, thank you.

There are no Flags listed at all

(except while I reboot one of the controllers, since then are not enough licenses for all APs concurrently)

I wonder why @ the failing  APs in 'show ap database' under 'Switch IP'  not the aruba-master IP addr,

rather the backup controllers IP addr are listed

check the LMS-IP in the AP system profile of that ap-group and make sure it does not point to an ip address that does not exist.  your access point could be reaching the backup controller via DNS, or DHCP options, and your LMS-IP could be sending them off to la-la land if it does not point to an actual controller.

Did you remove the AP from controller? ...  then try to purge the AP..

connect the AP again and wait until controller apply config..

From the CLI, do the command "clear gap-db ap-name xxx", where xxx is the name of the down AP. Do this for each AP you want to remove.

GUI - click on Monitoring > All Access Points > Clear (to the right of the AP you want removed).

Regasrds,

Giovanni Saldarriaga

Thank you for the responses and hints:
Yes, the APs  were power cycled /rebooted (even controllers) .

If a in 'show ap database' down listed AP IP addr is pinged from the controller# prompt, it responds,

but the following response to

controller#apboot ip-addr 10.10.10.200
AP is down

I shall try now your hint, also only see the possibility  to clear and reconfigure the [inaccessible?] APs from scratch

(btw There are enough free licenses available)

I wonder why @ the failing  APs in 'show ap database' under 'Switch IP'  not the aruba-master IP addr,

rather the backup controllers IP addr are listed

---

@gsaldarriaga@cosma.com wrote:

Did you remove the AP from controller? ...  then try to purge the AP..

 

connect the AP again and wait until controller apply config..

 

From the CLI, do the command "clear gap-db ap-name xxx", where xxx is the name of the down AP. Do this for each AP you want to remove.

GUI - click on Monitoring > All Access Points > Clear (to the right of the AP you want removed).

 

Regasrds,

 

 

Giovanni Saldarriaga

---

Thank you Giovanni,  your hints helped to clear the situation!

1) show ap database | include Down

2)  "clear gap-db ap-name xxx"

3) Power cycle each AP  *

AP does reconnect to a controller

4) show ap database | include xxx

If it reconnected  to the lms-backup  controller

5)    apboot ap-name

Else if connected to the home- / aruba-master controller directly

all APs are working properly again:

6) sh ap active

* rather difficult if the APs at all the remote sites are attached to an non-PoE switch...

Now I am searching for a) the cause,

and b) why it did not switch back from the lms-backup to the main- /aruba-master controller

@

ap system-profile <profile>
lms-preemption
lms-hold-down period <seconds>

I have no   "lms-hold-down period"  ,  normally it does /did work did failback 

do you know what is the default time in <seconds> ?

@ cjoseph  thanks,  the LMS-IP setup is correct