Gentlemen,

I see on the home web page that you are aware of the, as you state "Periodic Crashes".  I tell you that Periodic is NOT the case here in Ventura County.  I do not know about the Linksys systems but I can report that any and all of the Ubiquity systems that we have operational here are CRASHING sometimes several times a day.  Going to the mountain top to do a power reset because the node is NOT responding is getting pretty old.

We have been touting the technology and our Government official are asking for demonstrations.  We are reluctant to do this for fear of system crashes and we end up with looking bad.  At this point the system is so unreliable that we can NOT depend upon it for any meaning full communications.

Look, I know you people are all doing this as volunteers and out of the kindness of your hearts and is your way of advancing the technology, I commend you, BRAVO!!!  Right now we are dead as there is no parallel effort that I can see of people developing alternate code, we are depending entirly on your group.

If a FIX in NOT eminent, please give us a heads up as when to expect working firmware.  Please don't say soon, give us a date, two days, 5 days, Sept 15..Something so we are not left hanging here.  I do believe that great things can and will be accomplished with this technology and I will be the first to commend on your concepts and actions.

It was not my intent to make anybody in the design group mad or upset, if I was turse, please excuse me.  I do thank you for your time and listening.  We anxiously look forward to working software.

With great Respect

Paul Strauss

WD6EBY

Camarillo California

I was rummaging around inside a node the other night, and ran across quite a few NTPCLIENT files. I never did find a .conf file. I did find what looked like a perl script (maybe just a shell script) that included a NTPCLIENT call and a ping to the .pool ntp server address. It looked like it was intended to ping out to the world, and if it got a response, it would call the NTPCLIENT program, otherwise it would loop. It wasn't the olsrd.conf fiile, I'm pretty sure of that.  I'm sorry, I can't remember where I found it, but I found it by searching for NTP using WINSCP while connected to the node.

Just shooting in the dark, but perhaps the node won't look for ntp services if it doesn't see a path to a particular address? Perhaps if you could find that routine and plug in your servers' address...

I have to ask; what's the importance of setting the time on the nodes themselves? Aside from displaying on the admin pages, the node time doesn't seem to have any impact on the mesh. If I don't have a gateway up, my nodes usually show random times, and I haven't noticed any negative effects. I have to admit you've piqued my curiosity, but is there a practical consideration?

Hi Conrad,

I have NOT modified any of the code.  The only parameters that I have changed are those vi the WWW interface. 

BTW the number I left on your answering machine, call anytime from 9 am to 9 pm.

THX Conrad

Paul Strauss

WD6EBY

Output of 

 cat /sys/devices/pci0000:00/0000:00:00.0/subsystem_device

On the node please.

Have you modified any files on the node itself ? 

Hello,

I am sitting on the bench with several units talking to each other.  PicoStation2HP, Rocket M2, and a NanoStation loco M2.  Every thing is so close that I would expect and do read 100% under Mesh Status/Current Neighbors/ LQ.  Every thing except the loco unit it reads ~89%.  Don't fully understand.  I signed into the loco unit, under SETUP/BasicInformation/WiFi/active settings list and the the RX and TX antenna pull downs show up as "N Connector" and can not be changed.  I can see that this would hold true if I was working with a UBNT Bullet, but I am not, its a loco.

Yes I did do the proper factory flash with the proper software.  Does any body have any idea what is going on?

Thank you.

Paul Strauss

WD6EBY

There's a huge volume of stuff on Youtube. To wit:

https://www.youtube.com/playlist?list=PLDfWZ-E8CtRj0U61dbkC5fkk64-Iia4QC

It might take a while to get exactly what you need, but there's a lot of educational stuff.

At this time we are only able to generate images for the M series.

We currently are using more resources  (ram and flash) in the ar71xx build than are available on the previous hardware.

In addition OpenWrt has announced the devices with 16mb of ram are prone to overload in 12.09.

 Low resources is already a problem on Linksys builds (not enough flash and ram). The Linksys build already is unable to be brought to the same level Ubiquiti is at, the same issue would exist with pre-m Ubiquiti (less than 32/8)

Hi,

We have access to  Nanostation 2 but BBHN firmware is available only for Ar71xx. I have look on openwrt page :

http://wiki.openwrt.org/toh/start#ubiquiti

http://downloads.openwrt.org/snapshots/trunk/atheros/

and I have see that  UBT on atheros is supported by OpenWRT v8.09.

Is it possible generate BBHN firmware for "atheros" version ?

73 Waldek sp2ong

We only support the M series of Ubiquiti gear.

It needs to be a NanoStation M2 

Thank youthank you for all the information presented on the site because I debuted with Hamnet in the south of France.

I use a nanostation2 with the latest software version Ubiquiti.I can not seem to load the Hsmm by AirOS software file. A message says that the software is invalid.

I tested bbhn-1.1.2 and earlier version. Does anyone have a solution? 

Hello everyone,

Exactly.  I used to use Linux a lot and was always of the opinion that the "man" pages were wonderful if you KNEW how to do something.  If you didn't know, all you could do is hunt for an example from someone who had done it.

So, could someone, PLEASE, put up a few screen shots of the setup pages which are needed to set up services, port redirection, etc. 

Computers (usually) do what they are told to do.  But they are pedantic about the exact syntax of what they are  told.

PLEASE, PLEASE, PLEASE  a few EXAMPLES. 

73, Graham Lees ZL3GSL

I used Filezilla to do link speed testing this past weekend. Just install the app, identifiy a folder to share, then use any other computer from the same lan to test if FTP works. After that, it should be accessible from elsewhere on the mesh, using using only the IP address of the hosting computer.

You are using 5 node or 13 node direct on the hosting node aren't you? If you use NAT, you hide all of the devices on your node's LAN from all other users on the mesh unless you set up port forwarding. It certainly can be done, but so can driving to the store in reverse. One is lots easier than the other.

My Beaglebone Black hosts NTP, Asterisk, chat, a large file collection, get started web pages, and similar. The 5 node direct required no setup at all to see the device by name or IP, and only advertising of the services to let everyone on the same mesh see my devices and services.

Rick, NG5V

The post comes up a number of times about resetting passwords and/or changing the firmeware. Sometimes this is required to make sure the current load is valid, other times to return to factory firmware.

If you click the help link on several of the pages, you will find the answer to many of the frequently asked questions about what things mean and how to make changes such as this. The TFTP method and links to other tools like JTAG are found in the help section.

From any working mesh node, explore the help screens for a wealth of info. This is local to the device so anyone, anywhere can use the info.

Rick, NG5V

I had a friend give me his mesh nodes which are under his callsign and cannot remember the passwords.  I am not familiar with Linux that much, so if the solution requires that... please be patient.  How do I go about a re-flash?  I know the software needs updating, but I need to wipe it clean so I can restart with my callsign.

Thanks

Bill R.

N5VEI

The node's default is to go out a gateway to the specified ntp site. Without a gateway, the nodes don't seem to be able to sync up amongst themselves. I don't want to have to fiddle with every node's ntp config file to have them point to a specific node. That would require a manual setup for every deployment.

Just out of curiosity, could this be spoofed by adding an advertised service or named host with the ntp host named in the config file? ie, add a computer or gps time server to the mesh via a node, and give it a host name equivalent to the one the ntp clients are looking for? If a gateway isn't available, none of the nodes should be able to look at an internet DNS...

Think I just found next weekend's project.

Hmm. Re-read your post and think I missed the point. Sounds like your nodes aren't going out to get time updates, but devices connected to them are successfully updating. I haven't had that issue; every time I connect a node to a mesh with a gateway, the first thing that happens is it updates it's own time. I haven't updated yet, all my nodes are V 1.0. I've been interested in getting NTP working across a mesh without a gateway.