All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Nicolás Echániz" <nicoechaniz@codigosur.org>
To: The list for a Better Approach To Mobile Ad-hoc Networking
	<b.a.t.m.a.n@lists.open-mesh.org>
Subject: Re: [B.A.T.M.A.N.] Change metric and priorize links with the best Bandwidth
Date: Mon, 13 Aug 2012 22:09:46 -0300	[thread overview]
Message-ID: <5029A55A.7060500@codigosur.org> (raw)
In-Reply-To: <CAMVTf1gPjH6GiNy1fnoFCE3-qx42NBTZSgFafgTGBsaOW3tEcA@mail.gmail.com>

On 08/13/2012 08:18 PM, Esteban Municio wrote:
> Hi all
> 
> I'm asking myself if should be possible change the actual metric(TQ)
> in batman-adv or if there is a tool to configurte it.
> 
> We are deploying a rural mesh network where the nodes generally will
> be static in the same place, but maybe the links beetwen the nodes
> will have a unnestable bandwidth.
> 
> So we are looking for what protocol (with any plugin) is more
> efficient for our network.
> 
> Batman-adv have TQ like metric, but we would like that the nodes route
> the traffic priorizing the links with best bandwith available instead
> the links with best QoS. That is because finally, all the links tend
> to have ETX=1 at the expense of reduce de bandwidth(reducing the
> modulation and codification scheme in order to reduce looses), so the
> nodes route always for the shortest path...
> 
> Is there any solution for this?Do you recommend us change to another
> mesh protocol(OLSR+plugin) or another "static" protocol(ej: OSPF)? It
> is possible configurate batman-adv to get the fastest route instead
> the safer?
> Or maybe I am not understanding something...Am I proposing a fool thing?

At QuintanaLibre we experimented quite a bit with this and came to the
conclusion that setting a high multicast rate is a good solution to
"force" batman to only use good (in terms of bandwidth) links.

If you set the mcast_rate high, for example (OpenWRT):

config 'wifi-iface'
	option 'device' 'radio0'
	option 'mcast_rate' '54000'

then low bandwidth routes will effectively become invisible to batman.

We are using this setup in a semi-rural network, 3.5Km wide with 15
active nodes (as of today) and it works very well.

By doing this you loose redundancy, so if you have a node with only one
high bandwidth link and that link fails, the node won't have other
routes to use. In our experience this is OK because the alternative
low-bandwidth routes that are available if mcast_rate is very low are
not of much real use.


Cheers,
NicoEchániz

      reply	other threads:[~2012-08-14  1:09 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-08-13 23:18 [B.A.T.M.A.N.] Change metric and priorize links with the best Bandwidth Esteban Municio
2012-08-14  1:09 ` Nicolás Echániz [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5029A55A.7060500@codigosur.org \
    --to=nicoechaniz@codigosur.org \
    --cc=b.a.t.m.a.n@lists.open-mesh.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.