Veins - Mac1609_4 Problem

[Fabrício Silva]

Hi, 

when using the veins (http://veins.car2x.org/) example with other traffic model (different of the erlangen scenario provided), I'm getting the following error:

<!> Error in module (Mac1609_4) scenario.node[1579].nic.mac1609_4 (id=11066) at event #877979, t=29.404181564383: Model error: No packet was ready.

Does someone knows what is happening?

Thanks!

Fabricio

[Fabrício Silva]

For your information, I'm using the dataset TAPASCologne-sandesh.zip found in http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/

Fabricio

[David Eckhoff]

Hi Fabricio,

doesn't sound good. This means, that the Mac Layer wants to send a
packet but there is no ready packet in the respective EDCA Queue.

I cant reproduce the problem. Do you in some way interfere with the
packets after sending them to the mac layer? For example by deleting
them from the app layer?
I need more information to assist you.
Maybe you can try to create a minimal scenario based on the veins
example that causes this problem?

Greetings,
David

On 07/18/2012 06:57 PM, Fabr�cio Silva wrote:
> For your information, I'm using the dataset TAPASCologne-sandesh.zip

> <http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/TAPASCologne-sandesh.zip/download> found
> in http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/
>
> Fabricio
>
> Em quarta-feira, 18 de julho de 2012 13h42min58s UTC-3, Fabr�cio Silva

> escreveu:
>
> Hi,
>
> when using the veins (http://veins.car2x.org/) example with other
> traffic model (different of the erlangen scenario provided), I'm
> getting the following error:
>
> <!> Error in module (Mac1609_4) scenario.node[1579].nic.mac1609_4
> (id=11066) at event #877979, t=29.404181564383: Model error: No
> packet was ready.
>
> Does someone knows what is happening?
>
> Thanks!
>
> Fabricio
>

> --
> --
> Sent from the OMNeT++ mailing list. To configure your membership,
> visit http://groups.google.com/group/omnetpp
>
>
>


--
Dipl.-Inf. Univ. David Eckhoff
Computer Networks and Communication Systems
University of Erlangen, Germany
Phone: +49 9131 85-27627 / Fax: +49 9131 85-27409
mailto:eck...@cs.fau.de
http://www7.cs.fau.de/~eckhoff/

[Fabrício Silva]

Hi David,

I've downloaded the veins code (http://veins.car2x.org/download/veins-2.0-rc2.zip), and the erlangen scenario ran ok. 

Then, I've downloaded the koln.net.xml, koln_6-8.rou.xml and koln.sumocfg (TAPASCologne-sandesh.zip in http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/) and save these files into example/veins folder. I've also create the cologne.launchd.xml file, which is attached.

Then, I've changed the following in the omnetpp.ini:

sim-time-limit = 30000s

*.playgroundSizeX = 35000m

*.playgroundSizeY = 35000m

*.manager.launchConfig = xmldoc("cologne.launchd.xml")

and run the simulation and the error occurs:

<!> Error in module (Mac1609_4) scenario.node[1579].nic.mac1609_4 (id=11066) at event #877979, t=29.404181564383: Model error: No packet was ready.

I'm going to try to create a minimal scenario that causes the error.

Thanks!

Regards

Fabricio

Em quarta-feira, 18 de julho de 2012 14h59min05s UTC-3, David Eckhoff escreveu:

Hi Fabricio,

doesn't sound good. This means, that the Mac Layer wants to send a
packet but there is no ready packet in the respective EDCA Queue.

I cant reproduce the problem. Do you in some way interfere with the
packets after sending them to the mac layer? For example by deleting
them from the app layer?
I need more information to assist you.
Maybe you can try to create a minimal scenario based on the veins
example that causes this problem?

Greetings,
 David

On 07/18/2012 06:57 PM, Fabr�cio Silva wrote:
> For your information, I'm using the dataset TAPASCologne-sandesh.zip
> <http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/TAPASCologne-sandesh.zip/download> found
> in http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/
>
> Fabricio
>

> Em quarta-feira, 18 de julho de 2012 13h42min58s UTC-3, Fabr�cio Silva

[David Eckhoff]

Hi Fabricio,

Thank you very much, this should be enough.
If i have some time, i will look into this and report back.

Greetings,
David

On 07/18/2012 08:32 PM, Fabr�cio Silva wrote:
> Hi David,
>
> I've downloaded the veins code
> (http://veins.car2x.org/download/veins-2.0-rc2.zip), and the erlangen
> scenario ran ok.
>
> Then, I've downloaded the koln.net.xml, koln_6-8.rou.xml
> and koln.sumocfg (TAPASCologne-sandesh.zip

> <http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/TAPASCologne-sandesh.zip/download> in http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/
> <http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/>)

> On 07/18/2012 06:57 PM, Fabr�cio Silva wrote:
> > For your information, I'm using the dataset TAPASCologne-sandesh.zip
> >
> <http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/TAPASCologne-sandesh.zip/download
> <http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/TAPASCologne-sandesh.zip/download>>
> found
> > in
> http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/
> <http://sourceforge.net/projects/sumo/files/traffic_data/scenarios/TAPASCologne/>
>
> >
> > Fabricio
> >

> > Em quarta-feira, 18 de julho de 2012 13h42min58s UTC-3, Fabr�cio

> Silva
> > escreveu:
> >
> > Hi,
> >
> > when using the veins (http://veins.car2x.org/) example with other
> > traffic model (different of the erlangen scenario provided), I'm
> > getting the following error:
> >
> > <!> Error in module (Mac1609_4) scenario.node[1579].nic.mac1609_4
> > (id=11066) at event #877979, t=29.404181564383: Model error: No
> > packet was ready.
> >
> > Does someone knows what is happening?
> >
> > Thanks!
> >
> > Fabricio
> >
> > --
> > --
> > Sent from the OMNeT++ mailing list. To configure your membership,
> > visit http://groups.google.com/group/omnetpp
> <http://groups.google.com/group/omnetpp>
> >
> >
> >
>
>
> --
> Dipl.-Inf. Univ. David Eckhoff
> Computer Networks and Communication Systems
> University of Erlangen, Germany
> Phone: +49 9131 85-27627 / Fax: +49 9131 85-27409

> mailto:eck...@cs.fau.de <mailto:eck...@cs.fau.de>
> http://www7.cs.fau.de/~eckhoff/ <http://www7.cs.fau.de/~eckhoff/>

[Fabrício Silva]

Hi David, I've added some debug messages, and the output is the following:

** Event #877979  T=29.404181564383  scenario.node[1579].nic.mac1609_4 (Mac1609_4, id=11066), on selfmsg `{next Mac Event}' (cMessage, id=7900)

Channel turned busy: Switch or Self-Send

Stopping Contention at 29404181564383

Channel was idle for 0.000161999989

Passed slots after DIFS: 6

Updating backoff for Queue 1: 7 -> 0

Initiating transmit at 29.404181564383. I've been idle since 0.000161999989

Fabricio::iter->second.queue.size(): 0

Fabricio::iter->second.queue.size(): 1

Fabricio::iter->second.aifsn* SLOTLENGTH_11P + SIFS_11P: 7.1e-05

Fabricio:: iter->second.txOP: 0

Fabricio::iter->second.queue.size(): 0

Fabricio::iter->second.queue.size(): 0

<!> Error in module (Mac1609_4) scenario.node[1579].nic.mac1609_4 (id=11066) at event #877979, t=29.404181564383: Model error: No packet was ready.

So, the error is occurring in the method Mac1609_4::EDCA::initiateTransmit. It seems that the queue with size 1 has the txOP == false, and so it is not entering

the "if (idleTime >= iter->second.aifsn* SLOTLENGTH_11P + SIFS_11P && iter->second.txOP == true) {" condition.

Thanks again!

Fabricio

Em quarta-feira, 18 de julho de 2012 15h58min30s UTC-3, David Eckhoff escreveu:

Hi Fabricio,

Thank you very much, this should be enough.
If i have some time, i will look into this and report back.

Greetings,
 David

>     > Em quarta-feira, 18 de julho de 2012 13h42min58s UTC-3, Fabr�cio

[Fabrício Silva]

Hi again,

 I seems that the "iter->second.currentBackoff -= passedSlots;", in method Mac1609_4::EDCA::stopContent is setting the currentBackoff to zero, and

 then it is not entering the "if (iter->second.currentBackoff <= -1) {" condition.

Does this help?

Thanks

[Fabrício Silva]

Hi,

 I've cleaned the project and build it outside the Omnetpp IDE, and it is working fine. I could not find out what was going wrong...but it is working now.

Thanks

Fabricio

[TALBI Elhadi]

can you tell me from where this selfmsg provides ?
i've been searching for it in scenario.ned and didn't find the module sending this messages

Fabricio