Controllerless Networks

 

I am currently having issues with making an IAP225 a preferred master.

 

Before I join all the IAPs I usually pick one and make it the preferred master but when I bring up the rest of the IAPs the current preferred gets knock out by another master (non preferred) with default settings.

 

I get the following error:

Dec 31 16:00:54 udhcpc[2257]: Sending discover...

[   66.375382] (00:01:05) Fall into standalone, denied by hierarchy master (11:22:33:44:55:66 192.168.80.60)

 

Is there something I should be doing differently. 

 

 

IAP1c# [  715.506049] (00:11:55) Recover from asap standalone

[  715.564433] OOPS. someone else thinks hes the master too, beacon version 4

[  715.646753] I am MASTER. recv-ed a master normal-beacon.

[  715.710341] (00:11:55) master provision, 1 vs. 0

[  715.765579] (00:11:55) master provision, 1 vs. 0

[  715.820813] (00:11:55) !!! Election result, 1

[  716.360073] OOPS. someone else thinks hes the master too, beacon version 4

[  716.442422] I am MASTER. recv-ed a master normal-beacon.

[  716.506019] (00:11:56) master provision, 1 vs. 0

[  716.561252] (00:11:56) master provision, 1 vs. 0

[  716.616482] (00:11:56) !!! Election result, 1

[  717.360569] OOPS. someone else thinks hes the master too, beacon version 4

[  717.442895] I am MASTER. recv-ed a master normal-beacon.

[  717.506504] (00:11:57) master provision, 1 vs. 0

[  717.561734] (00:11:57) master provision, 1 vs. 0

[  717.616965] (00:11:57) !!! Election result, 1

[  718.360065] OOPS. someone else thinks hes the master too, beacon version 4

[  718.442426] I am MASTER. recv-ed a master normal-beacon.

[  718.506023] (00:11:58) master provision, 1 vs. 0

[  718.561255] (00:11:58) master provision, 1 vs. 0

[  718.616494] (00:11:58) !!! Election result, 1

[  719.360061] OOPS. someone else thinks hes the master too, beacon version 4

[  719.442408] I am MASTER. recv-ed a master normal-beacon.

[  719.506002] (00:11:59) master provision, 1 vs. 0

[  719.561236] (00:11:59) master provision, 1 vs. 0

Hi,

Have you tried to connect all the IAPs first in a cluster and then do the config on the VC. Once that is done the next step is to assign the preferred master while all the APs are online and see if that solves the problem. Also please note the VC IP needs to be out of the range of IAP IP address.

Hope this helps.

Thank you,

 

- What I usually do is that I reserve the VC IP address in DHCP

- I turn off PoE on all the ports but the IAP im planning to select as a Master 

- Upgrade if needed 

- Do the config

- And then turn PoE on the rest of the IAPs and this is where I run into an issue 

 

Is a hit or miss but it also depends on the amount of the IAPs in the cluster , I noticed that once I have more than 17 or 18 I see these issues but I am not 100% sure that this might be the issue.

Are you trying to do any of the following?

1. Deploying APs in hierarchical mode (where one AP is connecting to other APs through its downlink port)
2. Deploying LACP mode?

If neither of the above apply, could you check if you could have inadvertently plugged the ethernet cable into the E1 port instead of E0 port?

Thanks,

Yan

---

@Yan Liu wrote:
Are you trying to do any of the following?

1. Deploying APs in hierarchical mode (where one AP is connecting to other APs through its downlink port)
2. Deploying LACP mode?

If neither of the above apply, could you check if you could have inadvertently plugged the ethernet cable into the E1 port instead of E0 port?

Thanks,

Yan

---

- All the IAPs are currently connected to a S2500

- I have some of the IAPs configured using LACP and some are just using a trunk since there's not enough ports to configure all of the IAPs with LACP

- The ones using a single trunk should be using e0 but will need re-confirm that

 

 

If it is not too disruptive to the deployment, would you mind unplugging the E1 port from all APs to see if the problem is resolved? Then, assuming that the APs where E1 was used was for LACP, we need to debug to see why LACP did not work properly.

Yan Liu definately owe you a drink or several.

 

The issue was in fact because the IAP using a single trunk trying to take over was on e1 , it took me a little bit to find this out the actual IAP..

 

My other question will be if this is normal behavior ? because what would happen if with the other IAPs that are using LACP and e0 fails and then e1 takes over , would I see this issue again?

Hi Victor,

I am fairly confident that if LACP goes down this would not happen again.

Thanks,

Yan

Hi Guys , 

 

i have a set of 2 IAP105 that i am trying to add to a already connected iap225*2 for a small office space . The problem is when i try adding these 105's they bring down the complete wireless . both the other iap225's restart and have to be powered off and powered on to get them back up ..

 

Note current setup 

2*225 IAP US = default vlan + eth0 connected only + no vc setup yet + default setup - plug and play + code 6.3.1.1-4.0.0.1 will upgrade to 6.4.0.2-4.1.0.0

 

want to add

2*105 IAP US + eth0 connected only  + code Version 6.4.0.2-4.1.0.0 

 

thanks .. also please suggest any best practices 

 

 

thanks ,