[B.A.T.M.A.N.] batman routing clarification

[B.A.T.M.A.N.] batman routing clarification

[Stefano Scipioni]

[-- Attachment #1: Type: text/plain, Size: 797 bytes --]

I am switching a 20 mesh node from olsrd but I have some problem in routing.
I have no gateway node because I use openvpn to have a tunnel with PC.
All nodes have:
  batman -g 0 -r 3 -o 3000 <interface>

node C has:
  batman -g 0 -r 3 -o 3000 -a 10.0.0.0/24 <interface>

I have several case like this:

A----good----B----good-----C-----10.0.0.0/24----PC
\---------------poor-------------/

three node can communicate with all others but radio link from A to C is
poor.

When node A want to communicate to node C it goes directly to C because
batman calculate a direct route to it.
But if I manually set routes to use node B connection from A to C is good.

What is my mistake?

-r parameter in batmand daemon works only in searching gateway node? ...or
it works every time, also with normal node?

[-- Attachment #2: Type: text/html, Size: 922 bytes --]

Re: [B.A.T.M.A.N.] batman routing clarification

[elektra]

Hi -

which version of batman are you using? And you should provide us with more 
details. Just saying Link is 'poor' or 'good' is very unprecise. Remember: A 
direct route with 6 Megabit performs similar to a two-hop link with 12 
Megabits on each link, since no retransmissions have to be made.

If you set the routes manually, it would be interesting to know the transfer 
performance of the following: 

A->C
A->B
A->B->C
C->A
C->B
C->B->A
 

How many OGMs does A count directly from C and via B? (And vice versa)
By the way: -o 3000 is quite slow for sending Originator messages if you have 
20 nodes only. The default should be fine.

cu elektra
 


> I am switching a 20 mesh node from olsrd but I have some problem in
> routing. I have no gateway node because I use openvpn to have a tunnel with
> PC. All nodes have:
>   batman -g 0 -r 3 -o 3000 <interface>
>
> node C has:
>   batman -g 0 -r 3 -o 3000 -a 10.0.0.0/24 <interface>
>
> I have several case like this:
>
> A----good----B----good-----C-----10.0.0.0/24----PC
> \---------------poor-------------/
>
> three node can communicate with all others but radio link from A to C is
> poor.
>
> When node A want to communicate to node C it goes directly to C because
> batman calculate a direct route to it.
> But if I manually set routes to use node B connection from A to C is good.
>
> What is my mistake?
>
> -r parameter in batmand daemon works only in searching gateway node? ...or
> it works every time, also with normal node?

Re: [B.A.T.M.A.N.] batman routing clarification

[a.anselmi]

Hi,

> which version of batman are you using? And you should provide us with more
> details. Just saying Link is 'poor' or 'good' is very unprecise. Remember:
> A direct route with 6 Megabit performs similar to a two-hop link with 12
> Megabits on each link, since no retransmissions have to be made.

Yes, this is true; anyway suppose this scenario

A----> B ----> C

where:  A --> B and B --> are 800 kbps links
and: A---->C is a 500 kbps link

Suppose we have to send 100 packtes form A to C

if the direct link A---->C is chosen we suppose a 100 packets transmission
while chosing B as router (repeater)  we would expect a 100+100 packets
transmission!
What above is well right using that bandwidht values (6Mb) but with
poor/lossy links we have also to consider the IP datagrams loss at network
layer ---> which will cause TCP retransmits at higher layers (think to
http connection) just based on that links.
For this reason maybe a kind of "bandwidth threshold" can help in
preferring a 500_Kb per link mulit-hop path rather than a 300_Kb link
direct path.

This is a my opinion elektra, no criticism... all what we want is only
give hints and tips in order to improve batman.
Ok?

--
Antonio

Re: [B.A.T.M.A.N.] batman routing clarification

[elektra]

> This is a my opinion elektra, no criticism... all what we want is only
> give hints and tips in order to improve batman.
> Ok?

Sure. Why do you have the impression that your hints and tips are not 
appreciated? 

cu elektra

Re: [B.A.T.M.A.N.] batman routing clarification

[elektra]

> What above is well right using that bandwidht values (6Mb) but with
> poor/lossy links we have also to consider the IP datagrams loss at network
> layer ---> which will cause TCP retransmits at higher layers (think to
> http connection) just based on that links.
> For this reason maybe a kind of "bandwidth threshold" can help in
> preferring a 500_Kb per link mulit-hop path rather than a 300_Kb link
> direct path.

Well, you certainly have a point here. The goal of a routing protocol should 
be to achive the maximum achievable bandwidth for all nodes using the 
network. There is a fancy word for this condition that one can use to brag 
around and sound like a scientist: equilibrium.

A route can consist of a large number of very good links - this would be one 
extreme. Or consist of a minimum of lousy links with maximum packet loss. 
The first option would consume a very high amount of available bandwidth since 
there will be many retransimissions disturbing others and it is likely to be 
relatively slow if your machines use one wireless interface only. The latter 
is likely to be completely unusable for anything more than transfering a few 
bytes.

The protocol should find the most effective compromise, of course.

With OLSR and ETX a two hop route with ETX=1,1 on each link won't be used if 
there is a direct link with ETX=2,1. The idea of ETX is to find the best 
compromise regarding loss and retransmissions. One has to note that this 
approach is only useful for wireless networks. ETX on wired connections 
sucks.

We are doing basically the same with Batman, but it is inherit in the 
algorithm. Usually a shorter route has a smaller propagation delay - so 
packets are approaching faster at the node that is counting. Only if the 
longer - and slower - route looses significant less packets Batman will 
choose the longer path. Batman has a small benefit on wired connections 
because they are usually much faster than wireless links and thus tend to be 
preferred.

In the previous example OLSR will prefer a single hop route with more than 50% 
packetloss, rather than the two-hop path with approx. 17% packetloss. 

There are some people suggesting that Batman preferres shorter routes too 
much, so it is interesting to collect such reports. But misconfigurations 
have to be ruled out when collecting such data. 

Something that could help to improve the situation as a workaround in this 
case is to increase the broadcast rate in the wireless settings of the cards. 
This will reduce the range of broadcasts and can improve the result, because 
the number of OGMs on stretched links will decrease.

cu elektra

Re: [B.A.T.M.A.N.] batman routing clarification

[a.anselmi]

Thank for reply
I just read it but maybe is late in the night or maybe I'm too tired...
but some statements are puzzling me. Sure tomorrow morning I'll have  a
more clean mind and take a look more deeper to your scripts.

ciao

Antonio


>> What above is well right using that bandwidht values (6Mb) but with
>> poor/lossy links we have also to consider the IP datagrams loss at
>> network
>> layer ---> which will cause TCP retransmits at higher layers (think to
>> http connection) just based on that links.
>> For this reason maybe a kind of "bandwidth threshold" can help in
>> preferring a 500_Kb per link mulit-hop path rather than a 300_Kb link
>> direct path.
>
> Well, you certainly have a point here. The goal of a routing protocol
> should
> be to achive the maximum achievable bandwidth for all nodes using the
> network. There is a fancy word for this condition that one can use to brag
> around and sound like a scientist: equilibrium.
>
> A route can consist of a large number of very good links - this would be
> one
> extreme. Or consist of a minimum of lousy links with maximum packet loss.
> The first option would consume a very high amount of available bandwidth
> since
> there will be many retransimissions disturbing others and it is likely to
> be
> relatively slow if your machines use one wireless interface only. The
> latter
> is likely to be completely unusable for anything more than transfering a
> few
> bytes.
>
> The protocol should find the most effective compromise, of course.
>
> With OLSR and ETX a two hop route with ETX=1,1 on each link won't be used
> if
> there is a direct link with ETX=2,1. The idea of ETX is to find the best
> compromise regarding loss and retransmissions. One has to note that this
> approach is only useful for wireless networks. ETX on wired connections
> sucks.
>
> We are doing basically the same with Batman, but it is inherit in the
> algorithm. Usually a shorter route has a smaller propagation delay - so
> packets are approaching faster at the node that is counting. Only if the
> longer - and slower - route looses significant less packets Batman will
> choose the longer path. Batman has a small benefit on wired connections
> because they are usually much faster than wireless links and thus tend to
> be
> preferred.
>
> In the previous example OLSR will prefer a single hop route with more than
> 50%
> packetloss, rather than the two-hop path with approx. 17% packetloss.
>
> There are some people suggesting that Batman preferres shorter routes too
> much, so it is interesting to collect such reports. But misconfigurations
> have to be ruled out when collecting such data.
>
> Something that could help to improve the situation as a workaround in this
> case is to increase the broadcast rate in the wireless settings of the
> cards.
> This will reduce the range of broadcasts and can improve the result,
> because
> the number of OGMs on stretched links will decrease.
>
> cu elektra
> _______________________________________________
> B.A.T.M.A.N mailing list
> B.A.T.M.A.N@open-mesh.net
> https://list.open-mesh.net/mm/listinfo/b.a.t.m.a.n
>

[B.A.T.M.A.N.] 919 build

[Michael Burmeister-Brown]

[-- Attachment #1: Type: text/plain, Size: 4201 bytes --]

Antonio,

I am continuing to have problems sshing into nodes.  Currently, it is a repeater.  Doing ifconfig eth1 (my wireless adapter) gives me this:

eth1      Link encap:Ethernet  HWaddr 00:19:D2:03:D2:00  
          inet addr:10.19.66.100  Bcast:10.19.66.127  Mask:255.255.255.128


I am associated with the public AP.

So I then do:

ssh 10.19.66.1 -l root
ssh: connect to host 10.19.66.1 port 22: No route to host

I never had problems to ssh into gateways or repeaters before.

What is wrong?  This really worked in prior builds...

Mike





-----Original Message-----
From: b.a.t.m.a.n-bounces@open-mesh.net on behalf of a.anselmi@oltrelinux.com
Sent: Wed 9/19/2007 2:12 PM
To: The list for a Better Approach To Mobile Ad-hoc Networking
Subject: Re: [B.A.T.M.A.N.] batman routing clarification
 
Thank for reply
I just read it but maybe is late in the night or maybe I'm too tired...
but some statements are puzzling me. Sure tomorrow morning I'll have  a
more clean mind and take a look more deeper to your scripts.

ciao

Antonio


>> What above is well right using that bandwidht values (6Mb) but with
>> poor/lossy links we have also to consider the IP datagrams loss at
>> network
>> layer ---> which will cause TCP retransmits at higher layers (think to
>> http connection) just based on that links.
>> For this reason maybe a kind of "bandwidth threshold" can help in
>> preferring a 500_Kb per link mulit-hop path rather than a 300_Kb link
>> direct path.
>
> Well, you certainly have a point here. The goal of a routing protocol
> should
> be to achive the maximum achievable bandwidth for all nodes using the
> network. There is a fancy word for this condition that one can use to brag
> around and sound like a scientist: equilibrium.
>
> A route can consist of a large number of very good links - this would be
> one
> extreme. Or consist of a minimum of lousy links with maximum packet loss.
> The first option would consume a very high amount of available bandwidth
> since
> there will be many retransimissions disturbing others and it is likely to
> be
> relatively slow if your machines use one wireless interface only. The
> latter
> is likely to be completely unusable for anything more than transfering a
> few
> bytes.
>
> The protocol should find the most effective compromise, of course.
>
> With OLSR and ETX a two hop route with ETX=1,1 on each link won't be used
> if
> there is a direct link with ETX=2,1. The idea of ETX is to find the best
> compromise regarding loss and retransmissions. One has to note that this
> approach is only useful for wireless networks. ETX on wired connections
> sucks.
>
> We are doing basically the same with Batman, but it is inherit in the
> algorithm. Usually a shorter route has a smaller propagation delay - so
> packets are approaching faster at the node that is counting. Only if the
> longer - and slower - route looses significant less packets Batman will
> choose the longer path. Batman has a small benefit on wired connections
> because they are usually much faster than wireless links and thus tend to
> be
> preferred.
>
> In the previous example OLSR will prefer a single hop route with more than
> 50%
> packetloss, rather than the two-hop path with approx. 17% packetloss.
>
> There are some people suggesting that Batman preferres shorter routes too
> much, so it is interesting to collect such reports. But misconfigurations
> have to be ruled out when collecting such data.
>
> Something that could help to improve the situation as a workaround in this
> case is to increase the broadcast rate in the wireless settings of the
> cards.
> This will reduce the range of broadcasts and can improve the result,
> because
> the number of OGMs on stretched links will decrease.
>
> cu elektra
> _______________________________________________
> B.A.T.M.A.N mailing list
> B.A.T.M.A.N@open-mesh.net
> https://list.open-mesh.net/mm/listinfo/b.a.t.m.a.n
>


_______________________________________________
B.A.T.M.A.N mailing list
B.A.T.M.A.N@open-mesh.net
https://list.open-mesh.net/mm/listinfo/b.a.t.m.a.n


[-- Attachment #2: winmail.dat --]
[-- Type: application/ms-tnef, Size: 5194 bytes --]

Re: [B.A.T.M.A.N.] batman routing clarification

[elektra]

Thanks!

One thing that just came to my mind: when you used -o 3000 it takes very long 
to fill the sliding window size of OGM statistics. A sliding window size of 
128 will take 384 seconds to fill up - more than 6 minutes! So if you just 
fire up the daemon and quickly start your measurements, the routing results 
can be quite arbitrary since there are not enough packets to make a 
reasonable decision.

But it might very well be the case that the broadcasts of the stretched link 
are propagating too far. In this case there is a workaround: You can increase 
the broadcast datarate if your wireless drivers/cards support this feature. 
This way you will decrease the range of OGM broadcasts. This has proven to be 
a good solution for mesh networks using OLSR at Freifunk.

Batman-0.2 still has some flaws in the routing metrics. The routing algorithm 
works nicely, but the metrics are sometimes unfortunate (choosing routes with 
bad performance in transmit direction, rumors saying it is preferring routes 
with small hopcount too much.) We are addressing this with 0.3 and in the 
experimental branch of batmand.

Also it could indicate a problem with the wireless devices.

cu elektra



> 2007/9/19, elektra <onelektra@gmx.net>:
> > Hi -
> >
> > which version of batman are you using? And you should provide us with
> > more
>
> batman 0.2
>
>
> details. Just saying Link is 'poor' or 'good' is very unprecise. Remember:
> A
>
> > direct route with 6 Megabit performs similar to a two-hop link with 12
> > Megabits on each link, since no retransmissions have to be made.
>
> poor respectly to radio signal, A to C without B give 100kbit/s unstable
>
> If you set the routes manually, it would be interesting to know the
> transfer
>
> > performance of the following:
>
> A->C: 100kbit
>
> > A->B: full
> > A->B->C: 1000kbit
> > C->A: 100kbi
> > C->B: 2500kbit
> > C->B->A: 1000kbit
> >
> >
> > How many OGMs does A count directly from C and via B? (And vice versa)
> > By the way: -o 3000 is quite slow for sending Originator messages if you
> > have
> > 20 nodes only. The default should be fine.
>
> Ok, I removed -o 3000 and changed antenna in A node (less sensitive to node
> C).
> Now default route for A is B, that it is correct.
>
> > > I am switching a 20 mesh node from olsrd but I have some problem in
> > > routing. I have no gateway node because I use openvpn to have a tunnel
> >
> > with
> >
> > > PC. All nodes have:
> > >   batman -g 0 -r 3 -o 3000 <interface>
> > >
> > > node C has:
> > >   batman -g 0 -r 3 -o 3000 -a 10.0.0.0/24 <interface>
> > >
> > > I have several case like this:
> > >
> > > A----good----B----good-----C-----10.0.0.0/24----PC
> > > \---------------poor-------------/
> > >
> > > three node can communicate with all others but radio link from A to C
> > > is poor.
> > >
> > > When node A want to communicate to node C it goes directly to C because
> > > batman calculate a direct route to it.
> > > But if I manually set routes to use node B connection from A to C is
> >
> > good.
> >
> > > What is my mistake?
> > >
> > > -r parameter in batmand daemon works only in searching gateway node?
> >
> > ...or
> >
> > > it works every time, also with normal node?