Looky1173
commented
10 years ago Imported from https://bugs.launchpad.net/or/+bug/1393039/comments/1
| Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Sat, 15 Nov 2014 16:58:28 GMT |
Closed Looky1173 closed 9 years ago
Looky1173
commented
10 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Sat, 15 Nov 2014 16:58:28 GMT |
Looky1173
commented
10 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Sat, 15 Nov 2014 19:32:39 GMT |
Another crash report which may help, attached.
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Mon, 24 Nov 2014 13:07:02 GMT |
I have just tried to use exp. version x2666 and a system crash occurred within a few minutes. Log file attached. Also attached is the NVidia GPU log showing the graphics core temperatures if this may help.
Dave
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Mon, 24 Nov 2014 13:08:05 GMT |
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Mon, 24 Nov 2014 17:40:47 GMT |
Another system crash, this time using the Cambrian Route.
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Mon, 24 Nov 2014 17:41:50 GMT |
Also attached is the NVidia monitor of GPU and CPU core temperatures.
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Mon, 24 Nov 2014 17:43:53 GMT |
I am in the process of changing my graphics card from a GTX 195 to a GTX 760. When I have received the new card and tried OR on it I will update you on whether or not this has made any difference to the system crash problem.
Dave
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | James Ross (twpol) |
| Date posted | Tue, 25 Nov 2014 14:13:58 GMT |
Looking at the Open Rails logs, it's stopping inside a Present() call - which is the application (Open Rails) asking the driver (NVIDIA) to display the next frame on screen. The hang exception means it has been stopped there for 60s. This definitely looks like a driver bug - have you checked you are using the latest NVIDIA drivers (and for other system components like CPU, motherboard)?
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Tue, 25 Nov 2014 15:28:51 GMT |
Hi James,
I am running driver version 340.52 which is the latest driver recommended by NVidia for the GTX 295 card.
Apart from monitoring CPU core temperatures, I have not checked the CPU or the motherboard out using any form of diagnostic tool yet.
I have loaded OR V0.9 though (from this website) and it runs perfectly for hours on end. If I immediately run the latest experimental version, it sometimes crashes within a few seconds. As I stated in my UKtrainsim post on the subject, it just feels like the latest experimental versions of OR put more load on the graphics card resulting in a crash. I discussed the problem with NVidia and they seem to think its a "weak card" which runs the display OK generally (browser, windows etc) but fails under the load of a video game.
I will know more when I receive the new GTX 760 I have ordered and tried it. In the meantime I will try and find some motherboard/CPU diagnostics to run.
Dave
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Tue, 25 Nov 2014 15:59:50 GMT |
Have just run Intel processor diagnostic test with a PASS result.
Now running Hot CPU tester Lite to try and test motherboard.
Dave
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Tue, 25 Nov 2014 16:58:03 GMT |
Hot CPU tester Lite ran for over 10 minutes with no errors, all cores running at 100% and temperature far higher than I have ever seen running OR, about 83 degrees F.
Downloaded Valley Graphics test and ran that for about 10 minutes, GPU core temperatures at about 73 degrees F with no failure.
Dave
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | David Martin Lewis (j72tankengine) |
| Date posted | Fri, 28 Nov 2014 17:22:47 GMT |
James,
I now believe the problem with system crashes was due to a corrupt Microsoft Net Framework. The diagnostics mentioned above included in their installation a reinstallation of Microsoft Net Framework and since running them, I have had no further problems. OR experimental versions now run as solid as a rock.
I have also now taken delivery of my new GTX 760 graphics card and will see if that works ok when I install it. I will then be able to run with the latest NVidia drivers also.
Sorry to have raised a false error report, will always check the OR supporting software is installed correctly in future.
Dave
Looky1173
commented
9 years ago | Property | Value |
|---|---|
| Posted by | James Ross (twpol) |
| Date posted | Mon, 01 Dec 2014 20:13:53 GMT |
Thank you for taking the time to investigate!
I am unsure what in the .NET Framework install could have caused such a problem, but I am glad it is resolved.
Imported from https://bugs.launchpad.net/bugs/1393039
Copied from my post in the UK TrainSim forums:
I have had a problem with OpenRails for several months now and have tried everything I can think of to solve it.
At random times, sometimes a few seconds into an activity, sometimes after an hour or so after starting an activity, my computer crashes with the message "no signal" on the display. Some railway sounds continue but there is no response from the horn (space) button, or any other buttons, even Ctrl-Alt and Delete. Nothing works and I have to force a power down and restart to do anything again.
This problem only ever happens when running OpenRails. TS2014 is unaffected as is everything else I do using Windows 7 64bit.
I have tried monitoring the GPU and CPU temperatures and they appear to be normal just before the computer crashes. I have taken out the graphics card, carefully cleaned the edge connector and the board generally. The computer is regularly cleaned inside and I have checked all fans work properly. I have a hefty power supply unit in the computer case so don't think lack of power is the problem.
I use the experimental versions of OR and have tried using older versions, tried reinstalling the support software again, have tried various permutations within the OR options menu but all to no avail. When OR does work it works beautifully, I get frame rates of 50 and upwards usually unless there is a lot of complexity in the particular part of the route being displayed.
I would be interested to know if anybody else has experienced this problem and if so, have they found a solution. I haven't raised this as a bug with OR launchpad as I feel the problem is too vague as yet, affecting any of the installed routes and any of the activities, but not all the time.
After various suggestions from the OR Trainsim community, I posted the following:
Hi Ged, Dennis and David,
Thanks for your input guys,
Ged, the "no signal" message is indeed a function of the monitor hardware, however I have tried disconnecting and reconnecting, and exchanging the monitor lead, (VGA by the way not digital) with no success. Upon reboot all works fine, with no disturbance to the connections needed. The sounds I hear are indeed OR rail sounds but they sound more like a repeating loop of about 1-2 seconds duration rather than a continuance of the actual train sounds from the activity. The computer does not respond in any way (sounds, lights etc) to keyboard or mouse input, and gives the impression of "nobody at home" with a hard reset the only apparant solution.
With regard to the OR log, I will attach it to a launchpad error report as I would do normally anyway, but I feel that this crash is so severe that OR does not get a chance to log anything useful in the last moments of operation.
Dennis, the last time OR ran totally reliably for me was Version 0.9, which is a very long time ago. The oldest experimental version available on James Ross's website is X2231 and this still displays the problem for me. If there are any copies of older exp. versions or indeed 0.9 itself, I would be grateful if you could let me know where they can be downloaded from. I would need the list of versions between V0.9 and X2231 to do a useful binary search.
David, I will submit an error report to OR but I have a feeling that the limited information I am able to provide at the moment will prove inadequate to their needs for a successful diagnosis of the problem.
I had a look at the Windows Event Log and could not see anything that referred to the OR application crashing. Most of the entries appeared to be concerned with the subsequent reboot.
Dave