Controllerless Networks

I have a cluster of 11 IAP-205-US's. I have 5 IAP-225-US's that will not join the cluster.

If fact, the 225's will not get (And keep) an IP from a DHCP server, so I cannot communicate with them at all, or at least not easily. If I put one of the problem AP's and a computer on the same VLAN, the computer ends up getting an IP in the subnet 172.31.98.0/23 (this must be being assigned form the AP/VirtualController). For example, last time I did that, my IP was 172.31.98.164. When I browse to https://172.31.98.1 (which is my computers DHCP assigned gateway), I am able to get on a VC that shows the one AP. Interestingly though, the AP has an IP of 169.254.0.0/16 (the default "I can't get an IP" subnet).

From the VC in the strage setup described above, I was able to update the firmware on that AP to the same version as those on my 11x 205 cluster (6.4.2.6-4.1.1.9). BUT, when resetting and putting the 225 back on the 205 cluster vlan, even if I "clear defualts" (which the VC/GUI says is equivelant to a factory reset), it does not join.

The only way I can talk to that AP again is to jump on a VLAN without a DHCP server, wait to get an address from the 172.31.98.0/23 subnet, and then go to https://172.31.98.1

 

I have not physically reset any of these five 225 AP's because they are mounted 30-40 ft up and my wiring guys have already left the job site with their extension ladders and lifts. The GUI says clearing the config is equivalent to a factory reset though.

Not sure it means anything, but we're using PoE, not PoE+.

My questions are:
-Was resetting the AP from the GUI the same as holding the reset button on boot?
-If not, is there another way to reset it completely without physically touching it?
-Is the 172.31.98.0/23 a standard subnet for Aruba, or is that clearly some remnant of someone else' config still residing on the AP/VC (one of the 225's was previously used, so that is my fear)
-If I set the AP to a "standalone" AP (as an available option in the VC/GUI) and then "clear the defaults", could that do any good? My thought with this is that it may kick it off whatever cluster it thinks its a part of)

#AP225
#AP205

Do you have the IAP205 and 225 using the same level of code?

 

Also make sure that the network on 225's isn't connected to E1 

By level of code, do you mean firmware? If so, they're all on 6.4.2.6-4.1.1.9

 

And using the E1 port would cause this issue? So I will need to physically look to see if they are connected to that port or not?

BTW, my install guys just let me know that they indeed plugged all the 225's into e1 :/ I'm going to head to the site with a lift and switch them all over. I'll let you know if that's the issue. Thanks!

When you use E1 it causes the AP to not form a cluster in a non-lacp setup.

That was it: our installers hooked all the 225's up to E1 instead of E0. Thanks for your quick response!

E1 only works or gets activated if POE+ is being supplied to the AP but E0 is the main uplink port to use to power the AP. E1 can be used for LACP in cases where it is needed.