Differences between revisions 30 and 31
Revision 30 as of 2019-09-25 18:42:31
Size: 5661
Comment:
Revision 31 as of 2021-10-23 16:45:47
Size: 5420
Comment: fixed bad power supply
Deletions are marked like this. Additions are marked like this.
Line 6: Line 6:
Currently Offline<<BR>> <<BR>>
Line 12: Line 12:
 * Alix 2D13
 * Ubiquiti Bullet M2 Hi Power

=== Software Installed ===

 * AlixCab
  * OpenWrt r35318
  * NoCatAuth
  * OpenVpn
  * Olsrd
  * Snmpd
 * Meraki MR24
Line 26: Line 16:
 * Host: rip (alix2 gateway)  * Host: rip (mr24)
Line 28: Line 18:
  * IPv6addr: 2001:470:e962:3e81::1/64
  
 * Host: mildew (ubiquiti bullet)
  * IPv4addr: 10.11.62.130/25
  * Channel: 1
  * SSID: www.personaltelco.net
  * BSSID: 00:15:6D:E6:AB:F3
Line 49: Line 31:
 * 2021-10-21: An outage started around June 7, apparently due to a failing power supply, replaced it and back online.

Node name: Communitecture (503-230-1293)
Live Date: April 29, 2006
Contact: RussellSenior


1421 SE Division Street
Portland, OR 97202
http://communitecture.net/

Equipment

  • Meraki MR24

Network Configuration

  • Host: rip (mr24)
    • IPv4addr: 10.11.62.129/25

Installers / Organizers

Maintenance and System Log

  • 2021-10-21: An outage started around June 7, apparently due to a failing power supply, replaced it and back online.
  • 2013-01-31: Updated alix2 to r35318 with linux kernel version 3.7.4 and IPv6. --RussellSenior

  • 2009-12-30: Today, with DonPark, worked on the node. We replaced the NuCab named "dryrot" with an AlixCab named "rip". Also tested a different PoE injector on "mildew" without success. Mildew, the metrix box on the roof, has been unstable for most of December. We purchased a Ubiquiti Bullet M2 HP about the time that our node benefactor and building owner died. Our intention was to get on the roof and replace the metrix, but that is on hold until the building access issues get sorted out by the heirs. In the meantime today, we plugged in the Bullet M2 to the AlixCab and with a directional antenna aimed it across the street to the east. I'm seeing some packet loss, perhaps on the DSL at the moment. --RussellSenior

  • 2009-03-26: Called because of a mis-wiring inside, had the upstairs on the node host's internal network. Added eth2 to our ptp network (br0), fixed up some crossover issues on their ethernet and now it's all fixed up. --RussellSenior

  • 2008-04-18: Got a report of an outage this afternoon. Rebooted the rooftop and the nucab. Arrived to find dhcpd wasn't running (or at least, it wasn't responding). Restarted dhcpd and it was back in business. --RussellSenior

  • 2007-10-24: Reflashed the rooftop metrix (mildew) with a pyramid-1.0b5-based image, basically a tweaked version of what we are currently using on Mississippi. Turned off the linkstat script (a missnet thing). We'll see if that improves things. --RussellSenior

  • 2007-10-18: Stopped by to help out with a router failure upstream from our gear. While there, took the opportunity to install a 5 port switch between the DSL modem and their new router, which gets us off of their internal network. We had talked about this a while back, and this was a good opportunity to do it. A call to Integra would probably get our port forwards installed too. We are still getting reports of outages there that seem to originate in the metrix. I see messages like "wifi0: stuck beacon; resetting (bmiss count 4)" and "eth0: increased tx threshold, txcfg 0xd0f01008". I really need to cobble up a new firmware image (done, but needs testing) and get it flashed (maybe tomorrow). --RussellSenior

  • 2007-08-31: Responding to reports of node outage, stopped to check. The /etc/udev/rules.d/z25_persistent-net.rules had a typo that resulted in interfaces getting the wrong names, and thus eth0 had gotten misplaced and access to the outside was lost. Corrected, rebooted, fixed. --RussellSenior

  • 2007-05-07: Responded to a call indicating the node was down. Checked the network cables, power cycled the roof in the process, node is back up. Restarted openvpn tunnel, can't ping 10.11.255.1 from a client. Not sure if that is expected or not. --RussellSenior

  • 2006-10-17: Responded to a call indicating the node was down. I failed to see the metrix wirelessly, and we rebooted it. Then I failed to get a DHCP lease. We eventually tracked it down to a bad cat5 cross-over between the nucab and and PoE injector. I installed one of my own. In the context of fixing our external access, Brian offered to pay for a switch to be placed upstream from the Communitecture network to isoloate it from the node, as I had discussed a few months ago with Brian Garrison.
  • 2006-05-08: Stopped by to set up port forwarding. Discovered the DSL modem is configured to do NAT on 192.168.1.0/24, which the D-Link is connecting to. So, just port-forwarding from the D-Link isn't going to work. Have to reschedule a time when my errors won't mess them up. I also had some trouble getting DHCP resolution on my laptop. Moved dryrot to a static IP on eth0 (192.168.0.10, the dlink's dhcp range starts at 100). --RussellSenior

  • 2006-04-29: We need to talk to Communitecture folks, Christine in particular, about integrating with their network. Currently we are plugged into their wireless router, using iptables to disallow our wireless traffic from connecting to their LAN. Our initial preference was to put the Nucab in front, hanging their network off of eth2, and allowing us to do traffic shaping as necessary to keep them happy. The downside of that is our nucab becomes a point of failure. In any case, we didn't have administrative access to their D-Link wireless router, so we've punted for now until we can work it out with Christine.
  • 2006-04-29: Node is live. - CalebPhillips

Notes

  • Install scheduled for April 29, 2006 starting at 10:30 am


[CategoryNode][CategoryPersonalTelco][CategoryNodeBehindNat]

NodeCommunitecture (last edited 2024-12-16 06:05:24 by RussellSenior)