hello all, i'll use a 5.2 ghz devices to connect 2 places. right now the computers run with 192.168.1.X ip addresses. what's the configuration that i should use on the AP, the SM and on the computers so i can work with the canopy devices?
1) Assign them addresses within the same IP block (i.e. 192.168.1.x) and you will be able to manage them
2) Add a secondary address to your workstation in the default IP block (i.e. 169.254.1.x).
Assuming multiple computers in the 192.168.1.x IP block and you want to reduce the risks of somebody other than yourself accessing the radios, I’d opt for #2 (e.g. assign the AP 169.254.1.10 and the SM 169.254.1.11 and give your workstation a secondary address of 169.254.1.5)
i understand that the AP and the SM works as a wifi network card and that i'm using the ether port of the computer, is this correct? so i'll only have one IP address for the Pc and one for the AP and one for the SM. with this scheme, how would you use the ip addresses?
i.e. Pc1 = 192.168.1.100 Pc2 = 192.168.1.105 AP = 192.168.1.1 SM = 192.168.1.5
taking as an example the 192.168.x.x subnet
this way they can communicate, both Pc's and the radios being inthe same subnet. am i correct with this?
'cause i read the system user guide and there it shows 2 ip addresses for the AP in 2 different subnets. and now that i read for the second time what you wrote, is this a right configuration?
Pc1 = 192.168.1.100 Pc2 = 192.168.1.105 AP = 169.254.1.10 / 192.168.1.1 (this is the 2nd ip addr of the AP) SM = 169.254.1.11
i believe that after reading a little bit more the manual this last configuration is the one to use, right?
Wrong. The LAN2 network configured on the AP is the "RF Private Interface". This network is only used by the AP to communicate with its SMs; the default is 192.168.101.1. Test it out like this:
TELNET from your PC to the AP. LOGIN to the AP with "root" and your configured password. TELNET from the AP to the SM using 192.168.101.2, where ".2" is the SM's SESSION ID. LOGIN to the SM with "root". RESET the SM as a test, if you want. or... EXIT back to the AP. EXIT back to your PC.
You cannot communicate from your PC directly to either the AP or SM using the AP's LAN2 interface.
The easiest way to explain it is to say that any and all Canopy infrastructure does not have to be on the same subnet as your computer IP addresses to function properly. Let me try to give an example.
Assume a network with:
(1) Access Point
(2) Subscriber Modules
You could do the following and the network would function:
Access Point LAN-1 IP: 192.168.1.100/24
SM #1 LAN-1 IP: 10.11.12.13/25
SM #2 LAN-1 IP: 169.254.107.99/30
Customer's IP connected to SM #1 - 10.100.1.5/24
Customer's IP connected to SM #2 - 10.100.1.6/24
The customer's IP addresses would obviously have to match up with however you have your network designed. But, the point is that the Canopy infrastructure can be on completely different subnets and they will still perform their duties, that is forwarding packets based on destination MAC address just like switches do. I like to think of each piece of Canopy equipment as a single port on a switch with a MAC address table that is capable of storing 4,096 MAC addresses.
Obviously in practice you should never do anything like this example I just gave, but I did it this way to illustrate my point. There are some other assumptions too, like NAT not being enabled on the SM's in this demo network, etc.
The simplest, but not necessarily the best, option would be to assign all host IP addresses – router, AP(LAN1), SMs, and PCs – in the same subnet. For example, the router could be 192.168.0.1, and each other host would be given an address of 192.168.0.x, a mask of 255.255.255.0 (24 bits, or “/24”), and a default gateway of 192.168.0.1. The PCs could have their addresses assigned by a DHCP server, but the AP and SMs would need static addresses outside the range assigned by DHCP. In this configuration every host can talk to every other host; every PC can ping, telnet, or browse to the AP and all SMs.
The customer’s PC, however, does not need to access the IP address of its SM to be able to get to the Internet via the AP and router. The customer’s PC only needs to get to the address of the router, not the SM and AP. If you want the customer to be able to access their local SM’s web pages, so you can step them through the diagnosis of a problem, then the above configuration might be appropriate. In this case, however, make sure every Canopy SM’s password is different; you don’t want one customer to access another customer’s SM.
What many ISPs do, and what other forum members have suggested in this thread and others, is this: You must assign customers’ PCs to the same subnet as the outbound router – 192.168.0.x to 192.168.0.1 as above – but assign your Canopy units to a different IP network, like 10.99.x.x (or 192.168.201.x or 172.19.123.x or whatever). You’ll also need to assign one of these 10.99.x.x “management” addresses to a PC in your office to access the Canopy units. Your Ethernet and Canopy networks then carry traffic for two separate IP networks; Ethernet switches and Canopy units (when NAT is disabled) operate at ISO layer-2, IP is layer-3; there is no conflict.
A Canopy network is just a big, wide-area Ethernet switch. If you attached four PCs to a single Ethernet hub or switch, assigned the mask 255.255.255.0 and the following addresses to the PCs:
The two 192.168.44.x PCs could ping each other, the two 192.168.55.x PCs could ping each other. The 192.168.44.x PCs, however, could not ping the 192.168.55.x PCs. A Canopy network works the same way.
If you’re new to both Canopy and IP networking, your Canopy network is small, and you may be imposing on your customers’ patience and typing skills to solve problems, you might want to stick with the first configuration above. When you become more comfortable with what works and what doesn’t, but certainly before your Canopy network grows to more than 50 APs and SMs, design a new IP addressing structure and switch everything over.