[announce] Afrimesh Network Management Dashboard - 0.9-r721 Ubuntu BETA

[Antoine van Gelder]

Building on the innovative research efforts and tireless support of
the Wireless Africa group at the Meraka Institute of the CSIR [1], the
sometimes terrifying growth of the Scarborough Wireless Users Group
[2] and tools developed by the Open Mesh team [3] I am pleased to be
able to finally announce the BETA release of the Afrimesh Network
Management Dashboard for Ubuntu Linux!

[1] http://wirelessafrica.meraka.org.za
[2] http://swug.za.net
[3] http://www.open-mesh.net

Featuring:

* Powerful mesh network routing with B.A.T.M.A.N.
* Dynamic GIS visualization of your mesh on OpenStreetMap maps.
* Plan your network using the terrain elevation map.
* Visualize live health&traffic information for any mesh node.
* Keep inventory of your network devices.
* Monitor the health&usage of your Internet gateway.
* Monitor live network accounting information from pmacct.
* View network status & log messages in realtime.
* Customer management interface featuring FreeRADIUS support.
* Customer authentication featuring coova-chilli support.
* Online help and live chat support.
* Intelligent configuration assistance.
* Build and manage mesh networks even without an Internet
connection.
* Now supports the VillageTelco Mesh Potato.

Installation instructions live here:

http://code.google.com/p/afrimesh/wiki/Installation

The impatient & brave can try:

sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 50CFBA3E
sudo sh -c "echo deb http://ppa.launchpad.net/afrimesh/ppa/ubuntu hardy main >> /etc/apt/sources.list"
sudo apt-get update
sudo apt-get install afrimesh-dashboard

Please request features and report your bugs:

http://code.google.com/p/afrimesh/issues

Help can usually be found on:

http://groups.google.com/group/afrimesh

#afrimesh on irc.freenode.net

The source is licensed under the New BSD License:

http://code.google.com/p/afrimesh

A live demo can be viewed at:

http://demo.afrimesh.org

Finally, this announcement would be incomplete without the ever
growing list of folk who have contributed to the development of
Afrimesh:

Chris Morris
Kobus Roux
Ajay Makan
George James
Johann Hugo
Sebastian Buettrich
Louis Coetzee
Dhiren Seetharam
Jonathan Mzengeza
Edwin Cowley
Cynthia Mlambo
Nkosinathi Msimango
David Ume
Steve Song
David Rowe
Elektra Aichele
David Carman

Thank you everyone!

[David Rowe]

Hi Antoine,

Congratulations on your release.

A couple of questions:

1/ How do you evaluate the link quality to plot the three link states?
Reason I ask is that I have a nice line of sight link to a node 200m
away with a batman score or 230-253. But there is evidence of
re-transmission so I wouldn't call it a perfect link:

PING 10.130.1.14 (10.130.1.14) 56(84) bytes of data.
64 bytes from 10.130.1.14: icmp_seq=1 ttl=64 time=2.10 ms
64 bytes from 10.130.1.14: icmp_seq=2 ttl=64 time=1.70 ms
64 bytes from 10.130.1.14: icmp_seq=3 ttl=64 time=1.53 ms
64 bytes from 10.130.1.14: icmp_seq=4 ttl=64 time=3.19 ms
64 bytes from 10.130.1.14: icmp_seq=5 ttl=64 time=2.46 ms
64 bytes from 10.130.1.14: icmp_seq=6 ttl=64 time=1.84 ms
64 bytes from 10.130.1.14: icmp_seq=7 ttl=64 time=4.54 ms
64 bytes from 10.130.1.14: icmp_seq=8 ttl=64 time=1.66 ms
64 bytes from 10.130.1.14: icmp_seq=9 ttl=64 time=2.13 ms
64 bytes from 10.130.1.14: icmp_seq=10 ttl=64 time=2.03 ms

This is probably a case of good signal strength but lots of
interference. How it impacts voice quality is TBD.

2/ When I hit the demo.afrimesh.org page there was a pause while it said
"loading maps". Was this loading a map from an open maps site onto your
server? Or was it loading something into my browser?

How would the "loading maps" stage work at a site with poor or no
Internet connectivity?

Thanks,

David

[David Carman]

well done dude!

d

> --
> You received this message because you are subscribed to the Google Groups "village-telco-dev" group.
> To post to this group, send email to village-...@googlegroups.com.
> To unsubscribe from this group, send email to village-telco-...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/village-telco-dev?hl=en.
>
>

[Antoine van Gelder]

On 26 Feb 2010, at 00:56 , David Rowe wrote:

> Hi Antoine,
>
> Congratulations on your release.

Tx David :)

> A couple of questions:
>
> 1/ How do you evaluate the link quality to plot the three link states?

At the moment I'm using a ye big ole thumbsuck based on the output of
the vis daemon & the performance of the routers under my desk.

Briefly:

{ router : "10.0.0.5", neighbour : "10.0.0.10", label : 1.00 },
{ router : "10.0.0.5", neighbour : "10.0.0.3", label : 1.09 },
{ router : "10.0.0.3", neighbour : "10.130.1.41", label : 1.00 },

There are four constants which I compare against the label field:

l_perfect -> currently @ 1.0 and sets route dark green
l_acceptable -> anything above this sets route light green
l_warn -> anything above this sets route orange
l_unusable -> anything above this does get displayed, but with
a very low alpha value so that weak routes don't
clutter the display.

I'm hoping for some real-world experiences to help set these to sensible
values.

Elektra, while we're on the topic, I'm a bit confused at the difference
between the B.A.T.M.A.N. 'score' numbers (0-255) and the 'label' from
the vis daemon. Is there a difference?

> Reason I ask is that I have a nice line of sight link to a node 200m
> away with a batman score or 230-253. But there is evidence of
> re-transmission so I wouldn't call it a perfect link:
>
> PING 10.130.1.14 (10.130.1.14) 56(84) bytes of data.
> 64 bytes from 10.130.1.14: icmp_seq=1 ttl=64 time=2.10 ms
> 64 bytes from 10.130.1.14: icmp_seq=2 ttl=64 time=1.70 ms
> 64 bytes from 10.130.1.14: icmp_seq=3 ttl=64 time=1.53 ms
> 64 bytes from 10.130.1.14: icmp_seq=4 ttl=64 time=3.19 ms
> 64 bytes from 10.130.1.14: icmp_seq=5 ttl=64 time=2.46 ms
> 64 bytes from 10.130.1.14: icmp_seq=6 ttl=64 time=1.84 ms
> 64 bytes from 10.130.1.14: icmp_seq=7 ttl=64 time=4.54 ms
> 64 bytes from 10.130.1.14: icmp_seq=8 ttl=64 time=1.66 ms
> 64 bytes from 10.130.1.14: icmp_seq=9 ttl=64 time=2.13 ms
> 64 bytes from 10.130.1.14: icmp_seq=10 ttl=64 time=2.03 ms
>
> This is probably a case of good signal strength but lots of
> interference. How it impacts voice quality is TBD.

Now that we have a basic display I think there's a lot of
interesting work we can do to integrate quality data from the routes
over time and build on real-world experiences with mesh VOIP.

I'm hopeful that we can get the link quality visualization to a point
where it reflects real-world performance closely enough that it can
be used as an aid to plan&tweak physical mesh layout.

If we got this right it could save a lot of effort climbing on roofs
and visiting of neighbors!

> 2/ When I hit the demo.afrimesh.org page there was a pause while it said
> "loading maps". Was this loading a map from an open maps site onto your
> server? Or was it loading something into my browser?

I'm curious, how long did you have to wait?

The bulk of the pause was probably made up from downloading the
OpenLayers javascript libraries over my poor 512/256 ADSL line!

As soon as it is loaded the "loading maps" display vanishes and is
replaced with the map tiles downloaded from the openstreetmaps.org
server.

In a real-world setup that file would have been located on your
dashboard server so you would probably not have had time to see
the message.

> How would the "loading maps" stage work at a site with poor or no
> Internet connectivity?

Essentially you'd have to generate the map data and then host it on
a server on your mesh.

Under Utility->Settings there is a field to set the map tile server.

We have scripts in various states of disrepair that help with this,
getting them ready for mass consumption is part of the work leading
up to RELEASE.

Given that generating map tiles involves doing evil things to
postgresql we may want to look at maybe hitting someone up
for resources to host a 'map-maker' service that does the
generation for you and presents a tarball for download or even
ships a DVD with the tiles.

Distributing the entire world is maybe not too practical (yet) as
it currently takes up around half a terabyte.

- antoine

[Steve Song]

Kudos Antoine! Have just installed it from the PPA archive without a
hitch. Looking forward to a little help with the visualisation server
though.

-S

--
Steve Song
Telecommunications Fellow, Shuttleworth Foundation

email: steve...@shuttleworthfoundation.org
work: +27 21 970 1220
mobile: +27 83 482 2088
skype: steve_l_song
blog: http://manypossibilities.net
twitter: stevesong