Thanks Punky, I'm not 100% sure what you mean - my system IS doing NAT for its various wireless clients, the actual Internet connection via a 3G USB device. That's the whole purpose of the project.
Are you saying that bridging wlan0 with (wired) eth0 in order to serve both worlds may be a problem? Or are you saying that the wireless drivers simply aren't reliable enough to use as an AP? In the latter case, do you think it would be a good idea to upgrade to kernel 2.6.34 from debian/experimental? Do you know if there other things in Voyage that will break with 2.6.34? And/or should I try the latest drivers from wireless.kernel.org? Appreciate your help! Norm On Fri, Jul 9, 2010 at 9:27 AM, Kim-man 'Punky' TSE <[email protected]> wrote: > Hi Norm, > > I myself would suggest to NAT'd wireless client instead of bridged access > point. > > I see ath9k is not at production quality at 2.6.30. > > Regards, > Punky > > On 7/7/2010 10:23 PM, Norman Henderson wrote: >> >> Hi, my problem looks similar to the following but it isn't clear if >> there is/was a solution: >> >> http://list.voyage.hk/pipermail/voyage-linux/2009-September/012094.html >> >> I'm using Voyage ONE - version 0.7 build date 20100603 on an Alix 3d3, >> as a combined FW (Shorewall), 802.11bgn AP (hostapd 0.6.9 with nl80211 >> - libnl 1.1, configured for wpa + wpa2 with psk), and file server >> (Samba). External connectivity is via an external 3G USB dongle. >> Eventually I hope to get the Asterisk stuff set up as well as printer >> sharing. >> >> Everything including the wifi works fine for an undetermined time, >> minutes or more likely a few hours. Then the radio appears to stop >> transmitting (disappears from the list of available networks on my >> Win7, WinXP and cellphone). >> >> Wlan0 is bridged to eth0 - since this IS my Net access, I've had to >> fall back to wired connectivity. I think the behavior was the same >> before they were bridged but I'm not certain. I get lots of messages >> in daemon.log that look like normal key exchanges etc. (to me - not >> experienced in this area), and then an unexplained "deauthenticated" >> message. If I stop and restart hostapd it all starts working again. >> >> Jul 5 16:48:38 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 IEEE >> 802.11: authenticated >> Jul 5 16:48:38 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 IEEE >> 802.11: associated (aid 1) >> Jul 5 16:48:38 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 RADIUS: >> starting accounting session 4C31F3A6-0000003C >> Jul 5 16:48:38 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 WPA: >> pairwise key handshake completed (RSN) >> Jul 5 16:48:40 voyage hostapd: wlan0: STA 00:13:02:be:7b:f9 IEEE >> 802.11: authenticated >> Jul 5 16:48:40 voyage hostapd: wlan0: STA 00:13:02:be:7b:f9 IEEE >> 802.11: associated (aid 2) >> Jul 5 16:48:40 voyage hostapd: wlan0: STA 00:13:02:be:7b:f9 RADIUS: >> starting accounting session 4C31F3A6-0000003D >> Jul 5 16:48:40 voyage hostapd: wlan0: STA 00:13:02:be:7b:f9 WPA: >> pairwise key handshake completed (RSN) >> Jul 5 16:48:42 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 IEEE >> 802.11: authenticated >> Jul 5 16:48:42 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 IEEE >> 802.11: associated (aid 1) >> ... lots of repetitions of the above ... >> Jul 5 16:48:44 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 IEEE >> 802.11: authenticated >> Jul 5 16:48:44 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 IEEE >> 802.11: associated (aid 1) >> Jul 5 16:48:45 voyage hostapd: wlan0: STA 00:1e:65:09:f8:66 IEEE >> 802.11: deauthenticated due to local deauth request >> >> There is no message saying the second station is also deauthenticated, >> but at this point the DNMA92 becomes invisible to the outside world, >> both Win stations are dropped and neither they nor the cellphone can >> see the "voyage" SSID any more. >> >> There are some rather unclear hints in the wireless.kernel.org lists >> that suggest this MIGHT be a known issue and MIGHT be resolved with >> their "bleeding edge" code - that doesn't excite me at all. Plus I >> would need to be using kernel 2.6.34, voyage ONE is using 2.6.30, and >> I don't know what the implications would be of upgrading just the >> kernel (e.g. from Debian's "experimental" repository). >> >> One more thing isn't clear - is there any utility or log that can tell >> me what the wireless network stack thinks it's doing or why it thinks >> there was a "local request" to deauthenticate? >> >> I can post config files, iw outputs, dmesg, etc. etc. if needed. Any >> help appreciated! >> >> Thanks, Norm >> >> _______________________________________________ >> Voyage-linux mailing list >> [email protected] >> http://list.voyage.hk/mailman/listinfo/voyage-linux >> > > > -- > Regards, > Punky > > Voyage Solutions (http://solution.voyage.hk) > * Embedded Solutions and Systems > - Mesh Networking, Captive Portal, IP Surveillance, VoIP/PBX > - Network Engineering, Development Platform and Consultation > > > > _______________________________________________ Voyage-linux mailing list [email protected] http://list.voyage.hk/mailman/listinfo/voyage-linux
