[isf-wifidog] WiFiDog - Do I Have To Hack the Access Point?

Mina Naguib webmaster at topfx.com
Lun 9 Mai 14:09:50 EDT 2005


On 9-May-05, at 1:44 PM, Bryan Gawronski wrote:

> Good afternoon,
>
> I am interested in the features WiFiDog offers but I have a bit of a 
> different setup.  Can you please tell me if WiFiDog will function with 
> this setup.
>
> We are looking to put 40 - 60 access points down a busy street in 
> Buffalo, NY.  The access points will talk with a single access point 
> back at the base station and then be routed through a captive portal 
> (WiFiDog?) to the internet.  A few access points will connect back to 
> a local store or museum it is attached to in order to use bandwidth at 
> that location.  What we hope to accomplish is have a single WiFi 
> network with three of four different captive portal gateways that all 
> talk to a single user database.  We need to authenticate users, 
> throttle bandwidth, and do some content filtering.
>
> We will not be able to hack our access points as the vendor is 
> donating them and supporting them (I don't think they would be happy 
> to find OpenWRT).  Can I run WiFiDog on Red Hat Linux machines as a 
> captive portal and leave the access points alone and still get the 
> functions?  In order to make WiFiDog work must you hack the access 
> point?  I understand that hacking the access point removes the need 
> for a Linux server but with 40 - 60 access points I don't mind have a 
> few Linux boxes sitting around.
>
> Thank you for your time.
>
> Bryan Gawronski

Hello Bryan

Thank you for your interest in WiFiDog.

The setup you describe sounds similar to our early hotspot 
installations.  Our setups consisted of a junk box running any flavor 
of linux (redhat/debian/whatever) with the captive portal software on 
it (at that time it was NoCat).  Attached to that box was the LAN 
interface, either with a PCI WIFI card, or an ethernet card connected 
to a regular accesspoint such as the linksys WRT54G.  During that time 
we configured the routers, without any re-flashing or hackery, to act 
as simple bridges (turn off routing decisions/dhcp/dns/etc . . .)

We then finished our software (WiFiDog) and replaced NoCat with WifiDog 
while maintaining the same 2-box setup.

Then we quickly got our act together and got rid of the stand-alone 
linux box, capitalizing on the existing linux environment in the 
WRT54G.

Even though I'm not sure how you're planning on getting the 60 APs to 
talk to one main one (wireless?/ethernet?), I don't think you'll have 
much trouble getting the setup you described going without hacking 
WiFiDog.  WiFiDog does not really care whether the interface you tell 
it is the "LAN" interface is actually a WIFI outlet or any other type 
of interface (ethernet/software VPN/etc.)





Plus d'informations sur la liste de diffusion WiFiDog