veins <!> Error in module (TraCIScenarioManagerLaunchd) scenario.manager (id=6)

[ware...@fau.edu]

Hello,

I am having the following error wen running a large VANET simulation in Veins:

veins <!> Error in module (TraCIScenarioManagerLaunchd) scenario.manager (id=6) at event #589896210, t=31598: std::bad_alloc: std::bad_alloc.

The stack had the following information:

No source available for "0x77090da8" 

veins [OMNeT++ Simulation]

veins.exe [7620]

Thread [5] 0 (Suspended : Container)

0x77090da8

tcl84!TclpGetDate() at 0x640d28

UnregisterBadMemoryNotification() at 0x76e88543

0x0

Thread [4] 0 (Suspended : Container)

Thread [3] 0 (Suspended : Container)

Thread [2] 0 (Suspended : Container)

Thread [1] 0 (Suspended : Signal : SIGSEGV:Segmentation fault)

gdb

I have heard this may be related to not sufficient RAM. I am using an 8 GB I7 machine. I am running OMNETPP from MINGW32 as this is a Windows 8 machine that apparently does not allow for running Veins directly.

I am attaching a picture with the simulation state at the moment of the crash.

I can avoid the problem by reducing the time of the simulation but I would appreciate any help to understand this problem

Thank you,

Wilmer Arellano

[ware...@fau.edu]

Hello,

I tried to run the simulation using a shorter time interval to prevent the error from happening but actually went into another error. The information bellow was collected from OMNETPP's stack and console. Help to solve this problem is appreciated.

Thank you,

veins [OMNeT++ Simulation]

veins.exe [7476]

Thread [5] 0 (Running : Container)

Thread [4] 0 (Running : Container)

Thread [3] 0 (Running : Container)

Thread [2] 0 (Running : Container)

Thread [1] 0 (Running : User Request)

gdb

<!> Error in module (TraCIScenarioManagerLaunchd) scenario.manager (id=6) at event #211653907, t=30394: Model error: ASSERT: condition count == activeVehicleCount false in function processVehicleSubscription, mobility/traci/TraCIScenarioManager.cc line 1078.

TRAPPING on the exception above, due to a debug-on-errors=true configuration option. Is your debugger ready?

[Christoph Sommer]

Hi Wilmer,

ware...@fau.edu wrote:
> Model error: ASSERT: condition count == activeVehicleCount false

I've also seen this error, in highly overloaded scenarios.

It seems like SUMO sometimes does not provide information about a
vehicle even though it was successfully added to the simulation, has
already departed, has not yet arrived, and is not currently teleporting.

The easiest workaround I can offer is to remove the ASSERT [1]. This
will have no other effect than Veins not updating the state of vehicles
that are unaccounted for. If I recall correctly, they will keep moving
in the same direction and at their last reported speed.


Best,

Christoph


[1]
https://github.com/sommer/mixim-sommer/blob/veins-2.0/src/modules/mobility/traci/TraCIScenarioManager.cc#L1078

--
Dr.-Ing. Christoph Sommer
Institute of Computer Science
University of Innsbruck, Austria
Phone: +43 512 507-53285 / Fax: -53079
http://ccs.uibk.ac.at/~sommer/

[ware...@fau.edu]

Thank you Dr. Sommer,

I tried the workaround but now I ran into the problem bellow. I would Highly appreciate help with this issue.

Thank you,

Wilmer

veins [OMNeT++ Simulation]

veins.exe [3524]

Thread [5] 0 (Suspended : Container)

0x77dd0da8

tcl84!TclpGetDate() at 0x670d28

UnregisterBadMemoryNotification() at 0x76988543

0x0

Thread [4] 0 (Suspended : Container)

Thread [3] 0 (Suspended : Container)

Thread [2] 0 (Suspended : Container)

Thread [1] 0 (Suspended : Signal : SIGSEGV:Segmentation fault)

gdb

[Christoph Sommer]

Dear Wilmer,

ware...@fau.edu wrote:
> I am having the following error wen running a large VANET simulation in
> Veins:
>

> <!> Error in module (TraCIScenarioManagerLaunchd) scenario.manager
> (id=6) at event #589896210, t=31598: std::bad_alloc: std::bad_alloc.
>

> I have heard this may be related to not sufficient RAM.

Yes, it looks like the simulation is using more memory than the machine
has. One possible cause would be a memory leak.

> GB I7 machine. I am running OMNETPP from MINGW32 as this is a Windows 8
> machine that apparently does not allow for running Veins directly.

I don't understand what you mean. As far as I know, MinGW is a port of
the GNU gcc and binutils to Windows. What do you mean by "not running
directly"?

Best,

Christoph

[ware...@fau.edu]

Dear Dr. Sommer,

After the first report I have been running the simulation on Windows 8 with OMNETPP 4.3, SUMO 0.15. For some reason, the first time I tried to run the simulation on Windows 8 it did not run, I thought it was Windows 8 related and decided to use MINGW32. Sorry I forgot to update.

1. How could I detect a memory leak? Task Manager does not show anything unusual.
   
2. Would this be something operating system related?

Thank you,

Wilmer Arellano

[ware...@fau.edu]

Problem solved.

1. Installed dual-boot Ubuntu 12.10, Windows 8 (a nightmare but worth it)
   
2. Upgraded to Veins 2.1, Sumo 17.1, and Omnetpp 4.3 on both OS. 
3. Windows 8 still shows the same problem.
   
4. The simulation runs smoothly on Ubuntu.

Wilmer

[foad hergholi]

[foad hergholi]

hi 

i simulate a scenario of vanet in omnet++, but when i run it there is a problem about it.how i can solve it:

the problem message is:

<!> Error in module (TraCIScenarioManagerLaunchd) scenario.manager 

(id=6) during network initialization: Model error: Could not connect 
to TraCI server.Error message: 10061: Unknown error. 

TRAPPING on the exception above, due to a debug-on-errors=true 
configuration option. Is your debugger ready? 

On Tuesday, December 17, 2013 5:10:13 PM UTC+3:30, foad hergholi wrote:

[foad hergholi]

hi 

i simulate a scenario of vanet in omnet++ on windows 7, but when i run it there is a problem about it.how i can solve it:

the problem message is:

<!> Error in module (TraCIScenarioManagerLaunchd) scenario.manager 

[foad hergholi]

hi 

i simulate a scenario of vanet in omnet++, but when i run it there is a problem about it.how i can solve it:

the problem message is:

<!> Error in module (TraCIScenarioManagerLaunchd) scenario.manager 

[sharad singh]

Dear sir

                    i am getting this error.

A runtime error occurred:

Attempted to read past end of byte buffer -- in module (veins::TraCIScenarioManagerLaunchd) helloNetwork.manager (id=6), at t=0s, event #1

Launch a debugger with the following command?

nemiver --attach=3217 &

please help me.my email id is shrd....@gmail.com