olsr-users - a mailinglist where users can post questions and discuss olsrd related issues.

headers Bhat, Ishwara | 2 Apr 2009 14:08

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users
Am Thursday 02 April 2009 14:08:49 schrieb Bhat, Ishwara:
> Hi,
>
> I need advice. I have simple 4 node 3 hop ad-hoc network.
Okay...
>
> With Hello = 1.0 and Hello validity= 5.0;  TC = 1.0, TC validity =6.0, I
> start OLSRd.
>
> How much time is required for the network to startup and fully discovers
> all the routes at each of the nodes?
This depends on....

1.) the routing metric (Hopcount will be "stable" faster than ETX)
2.) the loss rate on the links
3.) the mobility of the network.

Henning Rogge

*************************************************
Diplom Informatiker Henning Rogge
Forschungsgesellschaft für
Angewandte Naturwissenschaften e. V. (FGAN) 
Neuenahrer Str. 20, 53343 Wachtberg, Germany
Tel.: 0049 (0)228 9435-961
Fax: 0049 (0)228 9435-685
E-Mail: rogge <at> fgan.de
Web: www.fgan.de
************************************************
Sitz der Gesellschaft: Bonn
Registergericht: Amtsgericht Bonn VR 2530
Vorstand: Prof. Dr. rer. nat. Maurus Tacke (komm. Vors.), Prof. Dr.-Ing. 
Joachim Ender (Stellv.)

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users

Thanks. 
If none of the nodes is not mobile and if ETX is 1.0, is there any prediction for the discovery of all routes at
all nodes, as a function of Hello and TC time values?

Thanks,
Ishwar

-----Original Message-----
From: olsr-users-bounces <at> lists.olsr.org [mailto:olsr-users-bounces <at> lists.olsr.org] On Behalf Of
Henning Rogge
Sent: Friday, April 03, 2009 12:10 PM
To: olsr-users <at> lists.olsr.org
Subject: Re: [Olsr-users] Network stabilization time

Am Thursday 02 April 2009 14:08:49 schrieb Bhat, Ishwara:
> Hi,
>
> I need advice. I have simple 4 node 3 hop ad-hoc network.
Okay...
>
> With Hello = 1.0 and Hello validity= 5.0;  TC = 1.0, TC validity =6.0, I
> start OLSRd.
>
> How much time is required for the network to startup and fully discovers
> all the routes at each of the nodes?
This depends on....

1.) the routing metric (Hopcount will be "stable" faster than ETX)
2.) the loss rate on the links
3.) the mobility of the network.

Henning Rogge

*************************************************
Diplom Informatiker Henning Rogge
Forschungsgesellschaft für
Angewandte Naturwissenschaften e. V. (FGAN) 
Neuenahrer Str. 20, 53343 Wachtberg, Germany
Tel.: 0049 (0)228 9435-961
Fax: 0049 (0)228 9435-685
E-Mail: rogge <at> fgan.de
Web: www.fgan.de
************************************************
Sitz der Gesellschaft: Bonn
Registergericht: Amtsgericht Bonn VR 2530
Vorstand: Prof. Dr. rer. nat. Maurus Tacke (komm. Vors.), Prof. Dr.-Ing. 
Joachim Ender (Stellv.)

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users

Sorry about the TYPO in original email..

Please read " If none of the nodes is not mobile "  
as 
"If none of the nodes is mobile and if ETX is 1.0, what would be discovery time as a function of Hello and TC
timers" ?

Thanks,
Ishwar

-----Original Message-----
From: olsr-users-bounces <at> lists.olsr.org [mailto:olsr-users-bounces <at> lists.olsr.org] On Behalf Of
Bhat, Ishwara
Sent: Friday, April 03, 2009 2:39 PM
To: olsr-users <at> lists.olsr.org
Subject: Re: [Olsr-users] Network stabilization time

Thanks. 
If none of the nodes is not mobile and if ETX is 1.0, is there any prediction for the discovery of all routes at
all nodes, as a function of Hello and TC time values?

Thanks,
Ishwar

-----Original Message-----
From: olsr-users-bounces <at> lists.olsr.org [mailto:olsr-users-bounces <at> lists.olsr.org] On Behalf Of
Henning Rogge
Sent: Friday, April 03, 2009 12:10 PM
To: olsr-users <at> lists.olsr.org
Subject: Re: [Olsr-users] Network stabilization time

Am Thursday 02 April 2009 14:08:49 schrieb Bhat, Ishwara:
> Hi,
>
> I need advice. I have simple 4 node 3 hop ad-hoc network.
Okay...
>
> With Hello = 1.0 and Hello validity= 5.0;  TC = 1.0, TC validity =6.0, I
> start OLSRd.
>
> How much time is required for the network to startup and fully discovers
> all the routes at each of the nodes?
This depends on....

1.) the routing metric (Hopcount will be "stable" faster than ETX)
2.) the loss rate on the links
3.) the mobility of the network.

Henning Rogge

*************************************************
Diplom Informatiker Henning Rogge
Forschungsgesellschaft für
Angewandte Naturwissenschaften e. V. (FGAN) 
Neuenahrer Str. 20, 53343 Wachtberg, Germany
Tel.: 0049 (0)228 9435-961
Fax: 0049 (0)228 9435-685
E-Mail: rogge <at> fgan.de
Web: www.fgan.de
************************************************
Sitz der Gesellschaft: Bonn
Registergericht: Amtsgericht Bonn VR 2530
Vorstand: Prof. Dr. rer. nat. Maurus Tacke (komm. Vors.), Prof. Dr.-Ing. 
Joachim Ender (Stellv.)

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users

Hi,

We've just done a reasonably successful install of an OLSRd based  
meshing network with a mix of wired and 5GHz wireless links ( 20  
Access Points and 2 cabled links - running on 15 low power Debian  
Lenny based OLSRd routing nodes). So thank you to everyone involved in  
the project. We did all of the node installations using a script we  
wrote to create the node configurations from an XML description of the  
desired network layout and in due course we will open-source this  
script.

In principle all the nodes should be identical apart from which  
sub-net they serve, which IP(s) they are on and how many network  
interfaces they have - we have one node however which will simply not  
behave in a reliable manner. Sometimes it will acquire routes to all  
the other nodes and their sub-nets without problem, however at other  
times it will only route to its close neighbours (2 of them). All of  
the other nodes on the network show (correct) routes to the 'faulty'  
node and its sub-net and if I manually add a return route to the  
routing table of the faulty node then two way traffic can be sent  
between the nodes.

I'm using HTTP info plugin to monitor each node, but apart from the  
links and neighbours simply being missing from the 'faulty' node  
nothing else is obviously wrong i.e. the links that are present show a  
link cost of 1.000 and the neighbour present only shows 1 2-hop  
neighbour in its list.

Unfortunately its not even consistent, it took several unpluggings and  
re-pluggings of cables to get it to recognise the whole network when  
first installed, has run for a week and has now gone to showing only 2  
neighbours. It also intermittently shows 1 further neighbor! - I've  
tried restarting olsrd and rebooting the box (desperation set in...)  
and for the moment I've had to leave it with the most important route  
(to the DNS server) added statically to the routing table.

Does anyone have any suggestions as to how I go about diagnosing the  
problem further?

Thanks,

Alex

-- 
Technical Director - Zednax Limited
W: http://www.zednax.com
T: +44 333 444 0160 Ext. 1001
PLEASE NOTE: Our new number should be charged by your phone operator
at geographic (01/02) rates. Our old number will still work for the
forseeable future.
F: +44 161 660 8010

Zednax Limited is registered in England and Wales, Company no. 05321754.
Registered address: Meadow House, Meadow Lane, Nottingham, NG2 3HS.
Zednax Limited is VAT registered, VAT registration no. GB 855 4468 92.

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users

Am Friday 03 April 2009 11:41:12 schrieb Bhat, Ishwara:
> Sorry about the TYPO in original email..
>
> Please read " If none of the nodes is not mobile "
> as
> "If none of the nodes is mobile and if ETX is 1.0, what would be discovery
> time as a function of Hello and TC timers" ?
1. second:
- all nodes send a "empty Hello" message

2. second:
- all nodes can send a Hello with 1/2 asymmetric neighbor

3. second:
- all nodes can send a Hello with 1/2 symmetric neighbors
- all nodes send a TC with it's neighbors

So in theory the network could work after ~3-4 seconds...

Henning

*************************************************
Diplom Informatiker Henning Rogge
Forschungsgesellschaft für
Angewandte Naturwissenschaften e. V. (FGAN) 
Neuenahrer Str. 20, 53343 Wachtberg, Germany
Tel.: 0049 (0)228 9435-961
Fax: 0049 (0)228 9435-685
E-Mail: rogge <at> fgan.de
Web: www.fgan.de
************************************************
Sitz der Gesellschaft: Bonn
Registergericht: Amtsgericht Bonn VR 2530
Vorstand: Prof. Dr. rer. nat. Maurus Tacke (komm. Vors.), Prof. Dr.-Ing. 
Joachim Ender (Stellv.)

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users
Am Friday 03 April 2009 12:17:23 schrieb Alex Masidlover:
> Hi,
>
> We've just done a reasonably successful install of an OLSRd based
> meshing network with a mix of wired and 5GHz wireless links ( 20
> Access Points and 2 cabled links - running on 15 low power Debian
> Lenny based OLSRd routing nodes). So thank you to everyone involved in
> the project. We did all of the node installations using a script we
> wrote to create the node configurations from an XML description of the
> desired network layout and in due course we will open-source this
> script.
>
> In principle all the nodes should be identical apart from which
> sub-net they serve, which IP(s) they are on and how many network
> interfaces they have - we have one node however which will simply not
> behave in a reliable manner. Sometimes it will acquire routes to all
> the other nodes and their sub-nets without problem, however at other
> times it will only route to its close neighbours (2 of them). All of
> the other nodes on the network show (correct) routes to the 'faulty'
> node and its sub-net and if I manually add a return route to the
> routing table of the faulty node then two way traffic can be sent
> between the nodes.
>
> I'm using HTTP info plugin to monitor each node, but apart from the
> links and neighbours simply being missing from the 'faulty' node
> nothing else is obviously wrong i.e. the links that are present show a
> link cost of 1.000 and the neighbour present only shows 1 2-hop
> neighbour in its list.
>
> Unfortunately its not even consistent, it took several unpluggings and
> re-pluggings of cables to get it to recognise the whole network when
> first installed, has run for a week and has now gone to showing only 2
> neighbours. It also intermittently shows 1 further neighbor! - I've
> tried restarting olsrd and rebooting the box (desperation set in...)
> and for the moment I've had to leave it with the most important route
> (to the DNS server) added statically to the routing table.
>
> Does anyone have any suggestions as to how I go about diagnosing the
> problem further?
What version of OLSRd you are using ?

Henning

*************************************************
Diplom Informatiker Henning Rogge
Forschungsgesellschaft für
Angewandte Naturwissenschaften e. V. (FGAN) 
Neuenahrer Str. 20, 53343 Wachtberg, Germany
Tel.: 0049 (0)228 9435-961
Fax: 0049 (0)228 9435-685
E-Mail: rogge <at> fgan.de
Web: www.fgan.de
************************************************
Sitz der Gesellschaft: Bonn
Registergericht: Amtsgericht Bonn VR 2530
Vorstand: Prof. Dr. rer. nat. Maurus Tacke (komm. Vors.), Prof. Dr.-Ing. 
Joachim Ender (Stellv.)

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users

Quoting Henning Rogge <rogge <at> fgan.de>:

> Am Friday 03 April 2009 12:17:23 schrieb Alex Masidlover:
>> Hi,
>>
>> We've just done a reasonably successful install of an OLSRd based
>> meshing network with a mix of wired and 5GHz wireless links ( 20
>> Access Points and 2 cabled links - running on 15 low power Debian
>> Lenny based OLSRd routing nodes). So thank you to everyone involved in
>> the project. We did all of the node installations using a script we
>> wrote to create the node configurations from an XML description of the
>> desired network layout and in due course we will open-source this
>> script.
>>
>> In principle all the nodes should be identical apart from which
>> sub-net they serve, which IP(s) they are on and how many network
>> interfaces they have - we have one node however which will simply not
>> behave in a reliable manner. Sometimes it will acquire routes to all
>> the other nodes and their sub-nets without problem, however at other
>> times it will only route to its close neighbours (2 of them). All of
>> the other nodes on the network show (correct) routes to the 'faulty'
>> node and its sub-net and if I manually add a return route to the
>> routing table of the faulty node then two way traffic can be sent
>> between the nodes.
>>
>> I'm using HTTP info plugin to monitor each node, but apart from the
>> links and neighbours simply being missing from the 'faulty' node
>> nothing else is obviously wrong i.e. the links that are present show a
>> link cost of 1.000 and the neighbour present only shows 1 2-hop
>> neighbour in its list.
>>
>> Unfortunately its not even consistent, it took several unpluggings and
>> re-pluggings of cables to get it to recognise the whole network when
>> first installed, has run for a week and has now gone to showing only 2
>> neighbours. It also intermittently shows 1 further neighbor! - I've
>> tried restarting olsrd and rebooting the box (desperation set in...)
>> and for the moment I've had to leave it with the most important route
>> (to the DNS server) added statically to the routing table.
>>
>> Does anyone have any suggestions as to how I go about diagnosing the
>> problem further?
> What version of OLSRd you are using ?

Version: olsr.org - 0.5.6-rc7 (built on 2008-07-12 17:23:53 on matrix.athome)

Thanks,

Alex

-- 
Technical Director - Zednax Limited
W: http://www.zednax.com
T: +44 333 444 0160 Ext. 1001
PLEASE NOTE: Our new number should be charged by your phone operator
at geographic (01/02) rates. Our old number will still work for the
forseeable future.
F: +44 161 660 8010

Zednax Limited is registered in England and Wales, Company no. 05321754.
Registered address: Meadow House, Meadow Lane, Nottingham, NG2 3HS.
Zednax Limited is VAT registered, VAT registration no. GB 855 4468 92.

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users

Am Friday 03 April 2009 12:28:29 schrieb Alex Masidlover:
> > What version of OLSRd you are using ?
>
> Version: olsr.org - 0.5.6-rc7 (built on 2008-07-12 17:23:53 on
> matrix.athome)
I would suggest to move to 0.5.6-r4.

Henning

*************************************************
Diplom Informatiker Henning Rogge
Forschungsgesellschaft für
Angewandte Naturwissenschaften e. V. (FGAN) 
Neuenahrer Str. 20, 53343 Wachtberg, Germany
Tel.: 0049 (0)228 9435-961
Fax: 0049 (0)228 9435-685
E-Mail: rogge <at> fgan.de
Web: www.fgan.de
************************************************
Sitz der Gesellschaft: Bonn
Registergericht: Amtsgericht Bonn VR 2530
Vorstand: Prof. Dr. rer. nat. Maurus Tacke (komm. Vors.), Prof. Dr.-Ing. 
Joachim Ender (Stellv.)

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users
> Am Friday 03 April 2009 12:28:29 schrieb Alex Masidlover:
>> > What version of OLSRd you are using ?
>>
>> Version: olsr.org - 0.5.6-rc7 (built on 2008-07-12 17:23:53 on
>> matrix.athome)
> I would suggest to move to 0.5.6-r4.

The latest .deb on the site is 0.5.6-r3 - would that be adequate to  
get to the bottom of the issue and I assume I'm going to have to  
install on all nodes and restart them all?

Thanks,

Alex

-- 
Technical Director - Zednax Limited
W: http://www.zednax.com
T: +44 333 444 0160 Ext. 1001
PLEASE NOTE: Our new number should be charged by your phone operator
at geographic (01/02) rates. Our old number will still work for the
forseeable future.
F: +44 161 660 8010

Zednax Limited is registered in England and Wales, Company no. 05321754.
Registered address: Meadow House, Meadow Lane, Nottingham, NG2 3HS.
Zednax Limited is VAT registered, VAT registration no. GB 855 4468 92.

--

-- 
Olsr-users mailing list
Olsr-users <at> lists.olsr.org
http://lists.olsr.org/mailman/listinfo/olsr-users

41	May 2013
32	April 2013
19	March 2013
52	February 2013
12	January 2013
62	December 2012
37	November 2012
23	October 2012
45	September 2012
31	August 2012
25	July 2012
17	June 2012
17	May 2012
7	April 2012
35	March 2012
67	February 2012
23	January 2012
16	November 2011
56	October 2011
10	September 2011
63	August 2011
9	July 2011
45	June 2011
101	May 2011
75	April 2011
98	March 2011
21	February 2011
93	January 2011
44	December 2010
67	November 2010
61	October 2010
87	September 2010
22	August 2010
49	July 2010
51	June 2010
43	May 2010
17	April 2010
40	March 2010
25	February 2010
107	January 2010
63	December 2009
58	November 2009
39	October 2009
63	September 2009
38	August 2009
18	July 2009
35	June 2009
43	May 2009
86	April 2009
27	March 2009
56	February 2009
141	January 2009
56	December 2008
51	November 2008
57	October 2008
126	September 2008
126	August 2008
53	July 2008
82	June 2008
88	May 2008
7	April 2008
45	March 2008
47	February 2008
24	January 2008
26	December 2007
82	November 2007
27	October 2007
17	September 2007
21	August 2007
23	July 2007
15	June 2007
39	May 2007
47	April 2007
21	March 2007
67	February 2007
65	January 2007
21	December 2006
15	November 2006
23	October 2006
23	September 2006
15	August 2006
62	July 2006
75	June 2006
41	May 2006
57	April 2006
78	March 2006
85	February 2006
28	January 2006
28	December 2005
48	November 2005
28	October 2005
3	September 2005
76	August 2005
80	July 2005
104	June 2005
67	May 2005
188	April 2005
54	March 2005
132	February 2005
93	January 2005
81	December 2004
58	November 2004
72	October 2004
25	September 2004

Network stabilization time

Re: Network stabilization time

Re: Network stabilization time

Re: Network stabilization time

A thank you and a question

Re: Network stabilization time

Re: A thank you and a question

Re: A thank you and a question

Re: A thank you and a question

Re: A thank you and a question

Mon	Tue	Wed	Thu	Fri	Sat	Sun

		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28	29	30