|
← Revision 11 as of 2009-01-08 03:21:40
Size: 4574
Comment: oops, missed a buffalo->netgear correction
|
← Revision 12 as of 2009-03-09 16:41:31
Size: 0
Comment: apparently no longer a node host
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 1: | Line 1: |
|
'''Node name:''' NodeIntegrity <<BR>> '''Live Date:''' September 6, 2007 <<BR>> '''Sponsor:''' Integrity Spirits <<BR>> '''Contact:''' info@personaltelco.net<<BR>> '''Website:''' http://www.integrityspirits.com/ 909 SE Yamhill St <<BR>> Portland, OR <<BR>> (503) 517-2030<<BR>> <<MapLink>> <<StreetView(sv?cbp=1,431.47988464492147,,0,7.417194166736491&cbll=45.515907,-122.656681)>> === Equipment === * Netgear WGT634U (Personal Telco's) * Netgear WPN824v3 === Software Installed === * OpenWrt Kamikaze r13753 * Wifidog === Network Configuration === * Netgear WGT634U * ESSID: www.personaltelco.net * Channel: 1 * BSSID: 00:0F:B5:0F:9B:DE * Hostname: integrity * Internet Network: 70.102.138.244/24 * Gateway: 70.102.138.1 * Upstream DNS servers: 204.130.255.3,64.122.32.71 * PTPnet Network: 10.11.5.1/26 * Netgear WPN824v3 * Connected via LAN ports to bridge to WGT * DHCP server disabled * ESSID: www.personaltelco.net * Channel: 6 * BSSID: 00:1E:2A:DE:59:48 * PTPnet IPaddr: 10.11.5.2/26 === Installers / Organizers === * RussellSenior === Maintenance and System Log === * 2009-01-06: Got a call today from Green Dragon again. Node outage. It has been failing and recovering for several days. I went to check. The buffalo could not talk upstream. Oddly, Integra could see the buffalo but not vice versa. We reset the dsl modem, which switched it back to bridging. But still, the buffalo could not talk upstream. At that point, Integra dude suggested trying another device, which was an excellent idea. My laptop worked fine. So, something was wrong with the buffalo. I took it home and returned a couple hours later with the old MISSNET-18 WGT634U, reflashed with a recent OpenWrt. Plugged that in, everything worked right away. Used some small screws to mount the wgt and dsl modem to the ceiling beam. Was at the Green Dragon for another hour chatting with DonPark and SethShikora who had shown up to play world of warcraft. Was still working (with wifidog turned on) when I left. 10 minutes after I left, around 9pm, Don called and told me he couldn't ping 10.11.5.1 (the wgt). He climbed up and power cycled it and it's been okay since. Not sure what happened. Watching. --RussellSenior * 2008-12-31: Got an email from Vince at Green Dragon that the internet was broken. They had integra and qwest out and they said it was fine. Turns out, they are plugged into a new DSL service. The DSL modem is now routing and the public IP address is now 70.102.138.244. They used to have a second DSL service. Maybe that went away and they just moved it??? Anyway, the Buffalo was configured with the old static IP and so it didn't work on the new line. So, we need to either get a port forward, or switch the DSL modem back to bridging and reconfigure the static IP. --RussellSenior * 2008-08-15: With DonPark, configured a Netgear obtained by the bar to bridge and configured its SSID to be consistent with the existing node. Moved the Netgear to near the Green Dragon bar and plugged it into a switched outlet (used to turn neon lights on/off). Should provide better coverage to outdoor area adjacent to QHUT. --RussellSenior * 2008-08-08: There have been numerous reports of problems with wifidog here. One of the problems seems to be: gee, i keep having to hit the splash page. The suggested solution from the wifidog dude was to force a short dhcp lease, the hypothesis being that firewall rules were dropping the icmp ping requests and preventing traffic from moving and thus wifidog sees the connection as absent and removes their auth-record. Today, we get a complaint from a user saying, geez your dhcp leases are awefully short. LuckyMike reports that apple laptops turn off their network when sleeping and thus get confused when they wake up again with no network. That might explain the 5 minute timeouts too, so I have up'd the number of intervals from 5 to 15 without traffic before dropping the auth-record, and also up'd the dhcp lease time to 14 minutes (from 4). We'll see how that affects things. --RussellSenior * 2007-09-06: In the evening, I returned to fix the firewall rules. The fix is to basically replace it all with a simple masquerade rule and call it good. --RussellSenior * 2007-09-06: Neglected to fix up the firewall rules, so remote ssh access is not working currently. --RussellSenior === Notes === * A distillery, bar and bistro. ---- [CategoryNode][CategoryPersonalTelco] |
- Diff for "NodeIntegrity"