[Opensim-users] OpenSim crashing why?

Mon May 3 19:21:52 UTC 2010

Hi

Justin wrote: "By the look of it, something in OpenSim is failing to
close connections.  After 24 hours, the HttpServer embedded in OpenSim
eventually finds that it runs out of sockets to play with."

OpenSim is still crashing periodically.

Any idea how I can find out how many connections the HTTP Server
embedded in OpenSim thinks it has open at any one time?

Is there a command from the Linux command line that will show this?
(I've tried ps and netstat without success)

Thanks very much

Clive

> Message: 2
> Date: Thu, 29 Apr 2010 23:38:49 +0100
> From: Justin Clark-Casey <jjustincc at googlemail.com>
> Cc: opensim-users at lists.berlios.de
> Subject: Re: [Opensim-users] OpenSim crashing why?
> Message-ID: <4BDA0A79.5020105 at googlemail.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Clive Gould wrote:
>> Hi
>>
>> We're successfully using OpenSim for student project work at Bromley College.
>>
>> OpenSim 0.6.8 standalone has just started crashing on us every 24 hours or so.
>>
>> Logins are disabled and the OpenSim.log file fills up the CentOS home
>> partition very quickly with error messages.
>>
>> Killing all processes owned by the opensim user, deleting the
>> OpenSim.log file, restarting the vnc session and the opensim
>> application cures the problem, until the next time.
>>
>> I append relevent excerpts from the error log to this message.
>>
>> Does anyone have any idea what may be causing this??
>
> By the look of it, something in OpenSim is failing to close connections.  After 24 hours, the HttpServer embedded in OpenSim eventually finds that it runs out of sockets to play with.
>
> I'm surprised to see this in standalone mode.  Are you doing something with OpenSim in standalone mode that requires outbound connections (e.g. remote database, groups, etc.), or perhaps a script that regularly establishes an outbound connection?
>
> If this has just started crashing after being fine before then it suggests that something has changed locally to reveal a bug in OpenSim.
>
> -- Justin
>
>>
>> Thanks
>>
>> Clive
>>
>> 2010-04-29 16:55:47,776 DEBUG - OpenSim.Region.Framework.Scenes.Scene
>> [Scene] Adding new agent India Airie to scene Rookery_NW
>> 2010-04-29 16:55:47,777 DEBUG - OpenSim.Region.Framework.Scenes.Scene
>> [APPEARANCE]: Appearance not found in Rookery_NW, returning default
>>
>> 2010-04-29 16:55:55,132 DEBUG -
>> OpenSim.Region.Framework.Scenes.SceneObjectGroup [SCENE]: Storing
>> Primitive, 204f5d0d-a994-4d61-8fb1-4b8d3af433fd in Rookery_
>> SW
>> 2010-04-29 16:56:14,262 DEBUG -
>> OpenSim.Region.Framework.Scenes.SceneObjectGroup [SCENE]: Storing
>> Primitive, 9b4c39ad-eebe-4417-9fd9-12a3994f17bc in Rookery_
>> SW
>> 2010-04-29 16:57:30,449 DEBUG -
>> OpenSim.Region.Framework.Scenes.SceneObjectGroup [SCENE]: Storing
>> Primitive, 88551bb8-3fa2-4bd9-a2fb-78c74bf45f55 in Rookery_
>> SW
>>
>> 2010-04-29 17:16:15,070 DEBUG -
>> OpenSim.Region.Framework.Scenes.SceneObjectGroup [SCENE]: Storing
>> Primitive, 9b4c39ad-eebe-4417-9fd9-12a3994f17bc in Rookery SW
>>
>> 2010-04-29 17:17:11,209 ERROR -
>> OpenSim.Framework.Servers.HttpServer.BaseHttpServer [BASE HTTP
>> SERVER]: HttpServer.HttpListener had an exception
>> System.Net.Sockets.SocketException: Too many open files
>>   at System.Net.Sockets.Socket.Accept () [0x00000]
>>   at System.Net.Sockets.Socket+Worker.Accept () [0x00000]
>> 2010-04-29 17:17:11,255 ERROR -
>> OpenSim.Framework.Servers.HttpServer.BaseHttpServer [BASE HTTP
>> SERVER]: HttpServer.HttpListener had an exception
>> System.Net.Sockets.SocketException: Too many open files
>>   at System.Net.Sockets.Socket.Accept () [0x00000]
>>   at System.Net.Sockets.Socket+Worker.Accept () [0x00000]
>> 2010-04-29 17:17:11,273 ERROR -
>> OpenSim.Framework.Servers.HttpServer.BaseHttpServer [BASE HTTP
>> SERVER]: HttpServer.HttpListener had an exception
>> System.Net.Sockets.SocketException: Too many open files
>>   at System.Net.Sockets.Socket.Accept () [0x00000]
>>   at System.Net.Sockets.Socket+Worker.Accept () [0x00000]
>> ockets.SocketException: Too many open files
>>   at System.Net.Sockets.Socket.Accept () [0x00000]
>>   at System.Net.Sockets.Socket+Worker.Accept () [0x00000]
>>
>
>
> --
> Justin Clark-Casey (justincc)
> http://justincc.org
> http://twitter.com/justincc
>
>
> ------------------------------
>
> Message: 3
> Date: Fri, 30 Apr 2010 08:08:45 +0200
> From: "Dirk Krause" <dirk.krause at pixelpark.com>
> To: <opensim-users at lists.berlios.de>
> Subject: Re: [Opensim-users] Looking for help reproducing a user study
>        inOpenSim
> Message-ID: <6C2AFDB53CBC0D4A8637CF3C7B63B2B8099F98 at postbox.bitlab.de>
> Content-Type: text/plain; charset="iso-8859-1"
>
> On a side note:  this is a perfect use case for an opensim-in-a-box (i.e. a cloud based temporary OpenSimulator with prdefined environment).
>
> just wanted to mention it.
>
> best,
>  dirk
>
>
> -----Urspr?ngliche Nachricht-----
> Von: opensim-users-bounces at lists.berlios.de im Auftrag von Elizabeth Lowry
> Gesendet: Do 29.04.2010 21:25
> An: opensim-users at lists.berlios.de
> Cc: Staff WorldSimTerra
> Betreff: Re: [Opensim-users] Looking for help reproducing a user study inOpenSim
>
> I can't personally run a reproduction of the study with participants I can't
> meet in person - I'm required to get consent forms from anyone who
> participates in *my* study.  That's why I'm looking for *other* people to
> run *reproductions* of my study, not for other places for myself to run my
> study.
>
> I don't need my own region, I already have my own grid!
>
> - Elizabeth
>
>
> On Thu, Apr 29, 2010 at 2:45 PM, Ideia Boa <ideiaboa at gmail.com> wrote:
>
>>  On 29-04-2010 8:31, Ideia Boa wrote:
>>
>> Hi Elizabeth
>>
>> We can build a region free of charge to all volunteers reproduce your
>> tests, and can also play one or two tests.
>> For other volunteers, you will have to be you to send invitations.
>>
>> The WorldSimTerra is always ready to host all kinds of initiatives for
>> schools and universities, only that, our time is very short so we can help
>> more.
>>
>> The region will have your name "ELIZABETH", so the volunteers will be easy
>> to find and may be available on Saturday and Sunday.
>> The owner of the region will be one of our mentors "Pato Donald", he will
>> have the ability to handle the server if necessary.
>>
>> Regards
>> Ideia Boa
>>
>> www.worldsimterra.com
>>
>>
>>
>> On 29-04-2010 8:06, Elizabeth Lowry wrote:
>>
>> Ideia,
>>
>>  Ah, then you can reproduce the study?
>>
>>  If so, that's wonderful!  Is there anything in the instructions I put up
>> at
>> http://hokiegrid.cs.vt.edu/index.php?&page=smodul&id=34&btn=33&subbtn=34 that
>> I should clarify to help you start?
>>
>>  Thanks,
>> Elizabeth
>>
>>
>> On Thu, Apr 29, 2010 at 1:44 PM, Ideia Boa <ideiaboa at gmail.com> wrote:
>>
>>> Hi Elizabeth
>>>
>>> We can build a region within our WorldSimTerra Grid for volunteers who can
>>> use it for your scientific work.
>>> Therefore all volunteers who are interested in reproducing your tests can
>>> do so in this region
>>> If there is interest in this region, please send a reply to the list and
>>> the call for volunteers.
>>>
>>> Regards
>>> Ideia Boa
>>> www.worldsimterra.com
>>>
>>> On 29-04-2010 9:15, Elizabeth Lowry wrote:
>>>
>>>  Hi all,
>>>
>>>  I tried telling the individuals who answered my first email back in
>>> March about this, but I haven't gotten any responses yet.
>>>
>>>  I'd really like to compare the results I'm getting from my study with
>>> someone else's - I've posted instructions at
>>> http://hokiegrid.cs.vt.edu/index.php?&page=smodul&id=34&btn=33&subbtn=34 and
>>> I'd really like to talk to some of you about reproducing my experiment.
>>>  Please?  It's for science!
>>>
>>>  Thank you,
>>> Elizabeth Lowry
>>>
>>>
>>> On Mon, Mar 22, 2010 at 10:29 AM, Elizabeth Lowry <emlowry at vt.edu> wrote:
>>>
>>>> Hi all,
>>>>
>>>>  My name is Elizabeth Lowry and I'm a Master's Student in the Computer
>>>> Science program at Virginia Tech.
>>>>
>>>>  As part of my thesis, I'm going to be running a user study using
>>>> OpenSim.  I'll be packaging my setup, including the data collection tools,
>>>> in an OAR, and I'm hoping some of you will be willing to reproduce my
>>>> experiment using that OAR and let me compare your results to mine.
>>>>
>>>>  The idea of the experiment is to measure the effect of levels of
>>>> spatial presentation in a virtual environment on that environment's
>>>> effectiveness.  I'll be presenting two lessons -- one on the photoelectric
>>>> effect, a topic with a level of inherent spatial content, and one on a
>>>> yet-to-be determined topic that isn't inherently spatial, like music or
>>>> logic -- at three levels of spatial presentation -- as a slide show given in
>>>> the virtual world, as a set of posters arranged in space (with the slides
>>>> from the previously-mentioned slideshow as the posters), and as a set of
>>>> full-on 3D exhibits (presenting the same information as the slides) taking
>>>> advantage of a virtual environment's ability to present 3D objects.
>>>>  Subjects will visit exactly one of the topic/level-of-spatial-presentation
>>>> combinations for a set period of time and their learning will be measured by
>>>> comparing their scores on quizzes before and after visiting.
>>>>
>>>>  The question I'm hoping to answer is: How much effort is really worth
>>>> it in the construction of a virtual learning environment?  And does this
>>>> level of effort vary between topics that are inherently spatial and ones
>>>> that aren't?  If the full-3D versions of the learning environments in the
>>>> study lead to the greatest improvements in quiz scores, it'll indicate that
>>>> the greater amount of effort required to build such environments pays off in
>>>> better learning.
>>>>
>>>>  Obviously, more work would need to be done in the future, whatever the
>>>> results of this study -- but to get solid results on this study, I'd like
>>>> your help.
>>>>
>>>>  Is anyone on this list in a position to run a user study in OpenSim
>>>> with 30 or more participants by mid to late April?
>>>>
>>>>  Thank you,
>>>> Elizabeth Lowry
>>>>
>>>
>>>
>>> _______________________________________________
>>> Opensim-users mailing list
>>> Opensim-users at lists.berlios.dehttps://lists.berlios.de/mailman/listinfo/opensim-users
>>>
>>>
>>> _______________________________________________
>>> Opensim-users mailing list
>>> Opensim-users at lists.berlios.de
>>> https://lists.berlios.de/mailman/listinfo/opensim-users
>>>
>>>
>>
>>
>> No virus found in this incoming message.
>> Checked by AVG - www.avg.com
>> Version: 9.0.814 / Virus Database: 271.1.1/2841 - Release Date: 04/28/10 20:27:00
>>
>>
>>
>>
>>
>> _______________________________________________
>> Opensim-users mailing list
>> Opensim-users at lists.berlios.de
>> https://lists.berlios.de/mailman/listinfo/opensim-users
>>
>>
>
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: not available
> Type: application/ms-tnef
> Size: 6068 bytes
> Desc: not available
> URL: <https://lists.berlios.de/pipermail/opensim-users/attachments/20100430/857020cb/attachment.bin>
>
> ------------------------------
>
> _______________________________________________
> Opensim-users mailing list
> Opensim-users at lists.berlios.de
> https://lists.berlios.de/mailman/listinfo/opensim-users
>
>
> End of Opensim-users Digest, Vol 32, Issue 35
> *********************************************
>