v2010 freeze with repeated runs

[I][the following thread has been edited by the moderator for brevity and attempted clarity due to the length of the exchange--just trying to make it easier to follow. Also, another thread was mixed in here regarding an independent issue, i.e., the RTRR function within DirectRT v2008--those posts were collected and are now located here: [URL]http://www.empirisoft.com/Support/showthread.php?t=5642[/URL]].[/I]

[I]___________________[/I]

[I][Sandra, see below][/I] and I are working together on a project and are having trouble with DirectRT v2010 freezing. It seems random. I am attaching the final csv file for you to look at. Hopefully you can tell me what is wrong. Would it have to do with the computer? Again, sometimes the file runs fine and collects data--and sometimes it just freezes.

What version of DirectRT and Windows are you are running it on?

[I][edited by moderator][/I]

We're running DirectRT v2010.122.804.10 on Vista and Windows 7. We tried to split the file, hoping that it would help -- it did not. We also increased the time interval to 1200.The freezing occurs sometimes at the beginning, sometimes the middle, and sometimes the end [...] It's now happening to a second experiment. Is it possible that there are automatic functions that the OS is performing that might be interfering? we've removed the virus scanning software and disabled the network. I'm attaching the two experiments (3 files).

I will try running these samples and see what I get. However I notice you are running a beta from august. Can you uninstall and try the release posted on the main download page (v2010.2.103 or later) and see if this still continues? See [URL="http://www.empirisoft.com/download.aspx"]www.empirisoft.com/download.aspx[/URL]. That may be all that's required. Again though, I will take a look at the samples you posted.

There is nothing immediately apparent to me in your input files that would cause any kind of crash. I would do one of two things. I've created simplified versions of one of your input files (experiment1.csv). One uses the exact same structure and stimuli on every trial--i.e., to rule out any kind of stimulus specific issue (e.g., see [I]experiment1[U]c[/U].csv[/I], attached). If this version produces no freezing, then it is likely a stimulus or experimental structure issue--interacting with some system or hardware issue. If the latter is the case, then please try running [I]experiment1[U]b[/U].csv[/I] which uses different stimuli as in the original but is [I]structurally simplified[/I] (it avoids all but the primary trial type).

If one or both of these test files can be run over and over again without a freeze it would take us a long way toward isolating the cause of the freeze.

Finally, I know you've done so, but do please double check to make sure any given machine is running DirectRT 2010.2.103 or later.

We tried DirectRT v2010.2.103 (the non-beta version) on the Windows 7 computer and had no problems with our first four participants, yesterday. Unfortunately, it froze during our session with the first participant, today.

[I]see reports from other users below[/I]

[I]see reports from other users below[/I]

Dave, how reliable is it that it will crash on the fourth run each time you restart? I did get a copy of your dxdiag file which is helpful. Can you also attach, here, a copy of your input file and style (.drt) file? Two other questions--what version do you have of DirectRT (i.e., 20xx.x.x) and is there a reliable point/time/trial or way in which it crashes during that fourth session?

Thanks Sandra. Was the computer rebooted or turned off at the end of the successful day? Or was it still in the state it was left in after the four subjects--i.e., when your fifth subject was run the next day? Also, is MediaLab involved concurrently with any of these sessions?

[I]see reports from other users below[/I]

Yes, we shut down at the end of the day [I][and so the system *was* rebooted before the subsequent first session the next day--i.e., the one that froze]. [/I]We now shut down between sessions and have have created multiple shorter sessions for each participant to minimze any data loss if one freezes. We also shut down after a freeze. We clear recent files as often as we can. We are not using MediaLab with these experiments.

I just modified a post above where I had proposed a test. Look for the post titled "trouble shooting test files". I just added a second test input file called [I]experiment1[U]c[/U].csv[/I]. It would be especially informative if you could run this file repeatedly as a test to see if it can cause the same problem.

And Bianca--in response to one of your earlier questions--yes, there is certainly the possibility of some parallel system event having an impact on this issue. Disconnecting from the network and shutting down any software or system services is a great start for trouble shooting. From your participants with incomplete data can you determine if there is any pattern in the time of day when the crashes occur? Maybe not to the minutes, but within certain hourly ranges? Or in terms of how long since the last reboot? If you'd like to send me a zipped folder with data from crashed and non-crashed sessions, I'd be happy to look over them for any kind of pattern I can find. That could speak to possible processes on your system that could be interfering. But again, a simpler way to start with that would be to simply shut everything down that you can aside from DirectRT. And also look for unusual running programs that all three machines have in common that perhaps are unique to your lab.

Finally, a last thought I have is that I could log onto one of your systems and take a look and see if I can spot anything unusual. Let me know and I can set up a remote support session.

Blair

Similar reports from other users:

From [I]spcl[/I], "Limited number of DirectRT sessions"
[quote]
I am using Direct RT 2010 on Windows 7. We have quite a few computers in our lab with the same setup. However, many of them cease to function after running a few experiments - at which point selecting any input file to run no longer produces anything, not even an error message. The same problem happens in the tutorial, where "Show me the session!" no longer works (although other buttons such as "Show me the input!" remain functional). Needless to say the "DirectRT Executable Mode" doesn't show up in the Task manager.

The problem disappears upon restarting the computer, but that's obviously time-consuming and repetitive. Please let me know what the problem is and how to fix it. Thanks!
[/quote]

From [I]DaveTTU[/I], "DirectRT Runs Input File Only Three Times"
[quote] I cannot run an input file more than three times without having to restart the computer. Is this because it is on Windows 7?
[/quote]

From [I]craigrl[/I],
[quote] I am having the same problem. Direct RT freezing at seemingly random points on three different computers. Using the beta version v2008.122.804.10 and running in Windows 7 (64 bit)... I am running an experiment using several DirectRT based IATs and am having trouble with it freezing at random intervals seemingly for no reason. I am running the IATs on four different laptops and they have frozen randomly several times on three of these laptops, so I can't imagine it is some fault with the computers themselves. I have also checked the input files which seem fine so don't think it is a problem with these (especially as the freezing occurs at different points each time it happens). When it freezes, DirectRT simply displays a blank screen, needs to be forced to close in task manager despite saying it is still running and loses all the data it has collected so far - simply replacing it with an empty folder called "voice".
Thanks,
Rebecca Craig
[/quote]

I was able to replicate a freeze here on one of our test machines. After a while I discovered there was an error message but it was located behind the session window. Presing ctrl+escape allowed it to appear. Can anybody let me know if they get a similar result?

[QUOTE]
This may offer a solution. I've a zipped and attached a revision of a file called executedrt.exe. The current version should be located here on your local machine:

[I]C:\DirectRT\Components\executedrt.exe[/I]

Try deleting (or re-naming) that file. Then unzip the attached revision (below) and use it to replace the older version.
[/QUOTE]

[I]This solution removed by moderator -- See below for alternate solution[/I]

I'm also running the DirectRT on Window 7, and from the day 1 I started using this program, it's been having the freezing problems (at least twice a day). And last week, I found this thread and downloaded the executedrt.exe file and replaced the older one.
But today, the program froze again - twice.

Apparently, the directrt has a compatibility issue with Window 7, so I wonder if I can resolve the problem if I run i from xp-mode (haven't tried but I heard that Window 7 allows me to run program from 'virtual xp'??)

Can you confirm the exact path (e.g., c:\...) of the current executedrt.exe file and right click on it and select properties to confirm its version? Do you happen to make a backup copy of the old version--can you check its version as well? Finally, assuming this all checks out, can you zip and attach copy of your experiment so I can try to replicate it here? And no, I would not recommend running it via emulated XP through Windows 7. If it's critical, I would install XP on the same machine in a dual boot configuration such that you can boot into XP or Windows 7. That way you can keep Windows 7 but have the option of booting up into a clean install of XP. Of course, that (at least for us) would only be a short term solution.

Both the original and the updated files were under
C:\DirectRT\Components. (I simply renamed the original as "executeDRToriginal" and kept it in the same folder)

And please find attached a copy of my test files.
But I run the same test on two computers: one is a newer computer and it has Window 7; another is a decade old computer and it has window xp.
So far, there's been no problem with the old computer, so I'm pretty sure there's nothing wrong with the test file.

Thanks for your help.

If any one posting on this thread could attach a dxdiag.txt report from their machine, it would be much appreciated. We have some ideas about why this is happening--just enter [I]dxdiag[/I] from the [I]Run [/I]command on the Windows [I]Start[/I] menu and save the results to a text file. We also have a newer test version that can be tried (2010.3).

Hope it helps!!!

I just realized that everytime the Voice folder was created the program froze. Although the Voice folder was NOT created EVERYTIME the program froze, I think if you could tell me how I can disable the program to create the voice log folder, it will help me a bit.

FYI, I have run 450 participants and the program froze 31 times. And the voice folders were created for 11 participants. None of those folders have a file. And there is no microphone connected to this computer.

We are having very similar problems, using 2010.2.103 on windows XP. Have already lost 25% of our data from random freezing. Has anyone found a solution to this problem? It seems to have something to do with the Voice Folder mentioned above. EVERY time it freezes, the folder is created along with an empty "Participant X" folder. Have tried the alternate execute file Jarvis wrote above, it seemed to work for awhile but just froze with my last participant. Frustrating...

Thanks

The gist is that we had some trouble working out compatibility issues for both MediaLab and DirectRT on Windows 7. This resulted in a few different approaches and a few very different versions.

For example, if your DirectRT folder has a "components" folder in it, then you have a version that *can* cause the issues described above. We've been working away from this set of releases. Currently DirectRT 2010.3 is the most compatible version we have available but it's not yet posted on the main download page (as of Feb 16, 2012). If you don't have it and would like to try it, contact our service@ empirisoft.com email address for a link to it. The primary trouble with v2010.3 though is that it will not work with most nVidia video cards. If you get an "overflow" error, that's almost always why. If you don't have an nVidia card, then v2010.3 might just fix you right up.

We have also just finished the first release of [B]v2012[/B] which seems, so far, to be the most compatible version of DirectRT yet, in terms of working with most releases of Windows (including Windows 7) and with the nVidia cards. It also has TTL compatibility in Windows 7 which was another snag on the previous versions.

Hi,

I'm having the same freezing problems described above using DirectRT v2012.4.145. One thing I noticed in the TrackPerformance text file was that I had 24 error messages that say either:

"Elapsed Time: 628439ms, block 0, trial 1011, stim 9, Error -2005532222, Automation error, on line 90, during CreateSurfaces of Form DDFullScreen"
OR
"Elapsed Time: 628521ms, block 0, trial 1011, stim 1, Error -2005532222, Automation error, on line 80, during FlipScreen of Form DDFullScreen"

These messages apparrently preceeded the freezing. Any idea as to why I'm experiencing this freezing?

Any help would be greatly appreciated. Thanks!

cforbes

Did you ever find a solution to this problem?