qosrns-general — Discussion about qosrns

Re: Fwd: About QoSR...

[Amardeep S. <ama...@ii...>]

> Hi sir,
>=20
> I've downloaded the source code of QOSR and tried the example but I saw=
 some
> problems.  I added the problems below...
>=20
> ns: _o10 create-session _o24 2.5Mb 1 _o151:
> =A0 =A0 (_o167 cmd line 1)
> =A0 =A0 invoked from within
> "_o167 cmd get-explicit-route _o24 2.5Mb"
> =A0 =A0 invoked from within
> "catch "$self cmd $args" ret"
> =A0 =A0 (procedure "_o167" line 2)
> =A0 =A0 (SplitObject unknown line 2)
> =A0 =A0 invoked from within
> "$rt get-explicit-route $dest $bw"
> =A0 =A0 (procedure "_o10" line 19)
> =A0 =A0 (Node create-session line 19)
> =A0 =A0 invoked from within
> "_o10 create-session _o24 2.5Mb 1 _o151" =A0 =A0 =A0
>=20

i saw your mail on ns mailing list. i think your problem has already been=
=20
solved.

--=20
Amardeep Singh

Fwd: About QoSR...

[Peng Z. <pg...@tc...>]

Hi Tarek,

I forwarded your question to the mailing list of the development group of=
 QoSR=20
in ns. Hope you can get quick response then.

Br,
Peng



----- Forwarded message from Tarek NADOUR <na...@in...> -----
Date: Mon, 6 May 2002 13:44:30 +0200
From: Tarek NADOUR <na...@in...>
Reply-To: na...@in...
Subject: About QoSR...
To: pg...@tc...

Hi sir,

I've downloaded the source code of QOSR and tried the example but I saw s=
ome
problems.  I added the problems below...

ns: _o10 create-session _o24 2.5Mb 1 _o151:
=A0 =A0 (_o167 cmd line 1)
=A0 =A0 invoked from within
"_o167 cmd get-explicit-route _o24 2.5Mb"
=A0 =A0 invoked from within
"catch "$self cmd $args" ret"
=A0 =A0 (procedure "_o167" line 2)
=A0 =A0 (SplitObject unknown line 2)
=A0 =A0 invoked from within
"$rt get-explicit-route $dest $bw"
=A0 =A0 (procedure "_o10" line 19)
=A0 =A0 (Node create-session line 19)
=A0 =A0 invoked from within
"_o10 create-session _o24 2.5Mb 1 _o151" =A0 =A0 =A0

can you make a sens out of this..

Thanks a lot in advance..

Tarek.


----- End forwarded message -----

About qosrns-0.2-prel

[Geunhyung K. <ge...@po...>]
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Re: Fwd: QoSR in Ns-2

[Amardeep S. <ama...@ii...>]

hi,

sorry for the late reply. my final exams were going on.

> ----- Forwarded message from Igor Canan=E9a (icc) <ic...@ci...> ---=
--
> Hi.
>=20
> I'm working on a network project at my university (UFPE - Brazil) using
> ns-2 and qosr 1.0.1.
> After creating examples and hacking through the code, I figured that th=
e
> path with the largest bandwidth available which suits my application is
> always the chosen one. But, I've made an example where that doesn't
> happen. I like to know if I wrong (maximum bw is not what counts).

actually it is not what is happening. the code is written to give a path=20
with shortest number of hops which has the required bandwidth available.
so, the script you posted works the way it is supposed to work.

On the other hand, it is easy enough to create a method which returns the=
=20
path with largest amount of bandwidth available. i think it should be eas=
y=20
to modify QOSRouting::get_od_explicit_route in new release of qosrns to=20
support that.

anyway, if you need some idea about how to create such a method, i=20
would be able to help you.

--=20
Amardeep Singh

Fwd: QoSR in Ns-2

[Peng Z. <pg...@tc...>]

Hi Igor,

I forward your email to the mailist of this development group, where you =
may=20
receive response more quickly.

Br,
Peng

----- Forwarded message from Igor Canan=E9a (icc) <ic...@ci...> -----
Date: Thu, 25 Apr 2002 01:42:43 -0300 (BRT)
From: Igor Canan=E9a (icc) <ic...@ci...>
Reply-To: Igor Canan=E9a (icc) <ic...@ci...>
Subject: QoSR in Ns-2
To: pg...@tc...

Hi.

I'm working on a network project at my university (UFPE - Brazil) using
ns-2 and qosr 1.0.1.
After creating examples and hacking through the code, I figured that the
path with the largest bandwidth available which suits my application is
always the chosen one. But, I've made an example where that doesn't
happen. I like to know if I wrong (maximum bw is not what counts).

I sending you the example in case you have a minute to look at it.
Hope you can help me.


	Igor Chaves Canan=E9a (ICC)
        Recife - PE


E-mails: ic...@ci...
	 igo...@gl...

-------------------------------------------------------------------------=
--
"In a world without fences and walls, who needs GATES and WINDOWS ??"

----- End forwarded message -----

Announce: qosrns-0.2-pre1

[Amardeep S. <ama...@ii...>]

Hi,

A new version of qosrns has been released.

It is a prelease of qosrns-0.2. I hope to release final version of it in 
a week. Final version would also ship with a version of mns for ns-2.1b8a 
with support for qosrns.

This prerelease can be found at
https://sourceforge.net/project/showfiles.php?group_id=35391&release_id=83927


Overview of changes in qosrns-0.2:
----------------------------------

1. qosrns is implemented as a seperate module, so it can coexist with
linkstate routing 

2. qosrns uses RtModule interface, thus making it easy to use it with
other routing modules

3. qosrns now supports on demand route computation. it is changed by
setting variable 
  Agent/QOSPF onDemand

4. qosrns now supports fine grained policy about when to send link
updates. it could be send periodically, as well as when bandwidth
changes by a certain percent. it can also use a hold down timer to not
send updates too quickly. it is controoled by modifying following 3
variables:
  Agent/QOSPF set advertInterval               	#periodic update interval
  Agent/QOSPF set advertHoldDownInterval       	#hold down timer for 
updates
  Agent/QOSPF set advertMinBwChangePercent     	#updates due to bw changes 
are only sent 
						when bw changes by this 
percent

--------------------------------------------------------------------------

-- 
Amardeep Singh

Re: help

[Amardeep S. <ama...@ii...>]

On Mon, 1 Apr 2002, Zuleika T. do Nascimento wrote:

> Sorry, I think I did not explain my problem clearly. 
> When I use create-session the method only measure between
> two  adjacent nodes (point to point),  e.g when the flow
> does not pass for  intermediate nodes. In addition the method
> generate the measurement data in a output file.
>  
> When I want measure end to end(passing among other intermediate nodes)
> the method doesnt generate errors , but also It doesnt store anything in
> output file .  
> Why does it happen ?
> Thanks !

but i cannot tell anything about it unless i know what method you are 
using to measure it. tell me something more about how are you measuring it 
then perhaps i could help you more.


-- 
Amardeep Singh

Re: help

[Zuleika T. do N. <zt...@ci...>]

On Fri, 29 Mar 2002, Amardeep Singh wrote:

> 
> 
> On Wed, 27 Mar 2002, Zuleika T. do Nascimento wrote:
> 
> > 
> > Hi,
> > 
> > Has anybody already developed a method to measure the delay and jitter
> > end to end,  starting the sources and creating session like the example
> > of qosr-simple.tcl available in patch of Qos Routing ?
> > 
> > I've implemented a method to do that , but it only works if I don't
> > create the session.
> 
> what problem occurs if you use create-session?
> 
Sorry, I think I did not explain my problem clearly. 
When I use create-session the method only measure between
two  adjacent nodes (point to point),  e.g when the flow
does not pass for  intermediate nodes. In addition the method
generate the measurement data in a output file.
 
When I want measure end to end(passing among other intermediate nodes)
the method doesnt generate errors , but also It doesnt store anything in
output file .  
Why does it happen ?
Thanks !

Zuleika.
> also, perhaps you can get a better response to this in ns mailing list
> 
> Amardeep Singh
> 
> 
> 

Re: help

[Amardeep S. <ama...@ii...>]

On Wed, 27 Mar 2002, Zuleika T. do Nascimento wrote:

> 
> Hi,
> 
> Has anybody already developed a method to measure the delay and jitter
> end to end,  starting the sources and creating session like the example
> of qosr-simple.tcl available in patch of Qos Routing ?
> 
> I've implemented a method to do that , but it only works if I don't
> create the session.

what problem occurs if you use create-session?

also, perhaps you can get a better response to this in ns mailing list

Amardeep Singh

help

[Zuleika T. do N. <zt...@ci...>]

Hi,

Has anybody already developed a method to measure the delay and jitter
end to end,  starting the sources and creating session like the example
of qosr-simple.tcl available in patch of Qos Routing ?

I've implemented a method to do that , but it only works if I don't
create the session.

If somebody has this method already implemented( with the creation of
session to measure the delay and jitter)  and could send me it
or could give some suggestion I 'll be very thanks.

Thanks for attention,

Zuleika.

Re: help me

[Amardeep S. <iam...@ya...>]

hi, 

firstly, i have created a mailing list for qosrns. you should subscribe
to it and post your queries there. you can subscribe to it by visiting
http://lists.sourceforge.net/lists/listinfo/qosrns-general

now, for your question

On Mon, 2002-03-11 at 17:53, Zuleika T. do Nascimento wrote: 
> 
> hi,
> 
> The method computerQOSroutes in the ls.cc program stores more the one
> combination (bandwidth,prevnodeid) associate a unique peer (destination,
> numhops). Is that affirmation correct? Because, I understood the
> algorithm only stores one combination, the last calculated, is it correct?

yes, it is correct.

> 
> Do you have a sugestion to modify the code source (ls.cc) to enable it to
> store all the combinations?
> 
> Thanks for attention.
> Zuleika.

there are two approaches to path computation - pre-computation or 
on demand computation. currently qosrns does pre-computation. you could 
modify the code such that path-computation is done on-demand. thatis, it should 
compute paths when asked for explicit route, such as in
list<int> LsRouting::get_explicit_route(int node, double bw)

there, you could copy the algorithm from computeqosroutes and modify it 
appropriately to get the path you need. in that scenario, there would
be no need to store all combinations. that would be imo much better option
in your case.

hope that helps. mail at qos...@li... if
you have further queries.

bye,

Amardeep


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com

(no subject)

[Antti <aer...@tc...>]

confirm 575513

Re: Qosrns-general -- confirmation of subscription -- request 575513

[Antti <aer...@tc...>]

qos...@li... wrote:
> 
> Qosrns-general -- confirmation of subscription -- request 575513
> 
> We have received a request from 130.233.154.41 for subscription of
> your email address, <ant...@tc...>, to the
> qos...@li... mailing list.  To confirm the
> request, please send a message to
> qos...@li..., and either:
> 
> - maintain the subject line as is (the reply's additional "Re:" is
> ok),
> 
> - or include the following line - and only the following line - in the
> message body:
> 
> confirm 575513
> 
> (Simply sending a 'reply' to this message should work from most email
> interfaces, since that usually leaves the subject line in the right
> form.)
> 
> If you do not wish to subscribe to this list, please simply disregard
> this message.  Send questions to
> qos...@li....