I have question regarding CAP setup on my AX2.
When I have in the CAP setup as “discovery interface” bridge where is ether1-5, than CAPsMAN is connected. But when I will change it only to ether2 where is router with CAPsMAN, than I have "no connection”, but when I add into CAPsMAN Addresses a IP of my router, connection is also working.
Question is why setup with only ether2 doesn’t work and what is correct setup.
But ether2 is inside the bridge?
Then it is not anymore a self-standing interface, it has been assimilated by the BORG bridge.
See this (good?) comparison:
normaly under bridge I have all interfaces. For me is hard to understand what’s wrong and what is correct setup.
My understanding is, that all ether must be under bridge to have data on all ether
My understanding in CAP configuration is, that I will set on which ether is CAPsMAN connected
Not when ether interfaces are part of bridge.
As indicated already (and to continue with the Borg analogy), if ether interfaces are part of bridge they have been assimilated by it so you need to use the upper level (bridge) as discovery interface.
When you use IP of controller, you use L3 (via bridge) and that's why it will work with that config.
But when it isn’t part of the bridge, data transfer isn’t working between ether which isn’t in the bridge and others which are in the bridge. Right ?
Question is, which setup is the correct one and which to use.
Discovery without any other parameter works on L2.
You can add IP address of controller but then you're moving to L3.
You can even add VLAN to the mix and then you need to be careful there the correct interface is specified.
So if it works or not, depends on your config.
Correct config for most AP running under capsman (without VLAN) is to have all interfaces connected to bridge.
There is a reason that's the default config when doing "reset to caps mode".
when I will use setup with ether2 and specified capsman IP address is can see on the CAPsMAN side :
and on the CAP side:
so I clearly see information about IP. When I will use a bridge, there is a information that it’s handled by bridge - so more confused
In order to hopefully un-confuse yourself:
The "right" or "correct" way is "discovery interface" on "bridge" (and all interfaces added to bridge).
Adding the IP of the router is possible but "wrong"[1].
[1] which does not mean that it cannot work, but the connection happens at a different level, L3 (IP) vs. L2 (MAC).
ok, then I will do it in that (bridge) way.
It just doesn't seem that clear to me:
in the bridge must be all ether and all wifi ?
Normally yes, the AP is a switch combining all the cabled interfaces and the radios.
If you prefer, most common AP's have only one ethernet port (the one with the connection to the "main" router) and one (or two) radio(s) all assembled in a bridge.
As well most common non-wifi routers have one "WAN" port (usually blue in colour) connected to the ISP/internet and (usually) 4 "LAN" ports (usually yellow in colour) assembled in a bridge, while wifi routers have the same arrangement with a "WAN" port and 4 "LAN" ports assembled in a bridge together with radio(s).
The above configurations are usually "fixed" and you buy either an AP or a router.
An Ax2 is a very flexible device, that allows to manage each interface by itself, but the most common configurations are the same as the above:
But you could do crazy things like (don't do it) make on it two bridges, one with all ether ports in it and one with the two radios, using the first bridge as a cabled switch and the two radios as an independent range extender, keeping the two bridges separated.
clear, thank you
