My CTD during MP

[topdog]

I get this after every couple of hours on average (but not every couple of hours mind you, it's still fairly erratic and out of the blue without warning).

 

I searched forums for the initial .dll reporting the problems (btmmhook.dll) and found no references (oddly I thought).

 

So here goes..

 

Patched to 1.0.2, removed all mods (so as to be integrity check compatible, though the only mods I ever had before were my own export.lua's for x52 pro shenanigans, and tacview).

 

Vista 32-bit, but don't let that freak you out ;)

 

One common theme is that it's always 2 crashes in series, though I guess that matters not in the grand scheme.

 

Copy/pasting the gubbins inside each of the windows that pop up, it has this to say. The first item is always the same, the second item is sometimes this .dll, sometimes kernel32.dll:

 

Problem signature:

Problem Event Name: APPCRASH

Application Name: simulator.exe

Application Version: 1.2.1.0

Application Timestamp: 4bed01e6

Fault Module Name: btmmhook.dll

Fault Module Version: 6.0.1.6300

Fault Module Timestamp: 47b20d5c

Exception Code: 40000015

Exception Offset: 0000e54f

OS Version: 6.0.6002.2.2.0.256.1

Locale ID: 2057

Additional Information 1: eefa

Additional Information 2: 7abdbd46d8dd56e61eadbb8242154000

Additional Information 3: 05d5

Additional Information 4: ccc05341b36b19cf0874468a911bef15

Read our privacy statement:

http://go.microsoft.com/fwlink/?linkid=50163&clcid=0x0409

 

Problem signature:

Problem Event Name: APPCRASH

Application Name: simulator.exe

Application Version: 1.2.1.0

Application Timestamp: 4bed01e6

Fault Module Name: MitkaGraphics.dll

Fault Module Version: 0.0.0.0

Fault Module Timestamp: 4becffee

Exception Code: c0000005

Exception Offset: 0002545a

OS Version: 6.0.6002.2.2.0.256.1

Locale ID: 2057

Additional Information 1: 843f

Additional Information 2: ac28f0976be6a2f8fcfab1cefe4fa2e5

Additional Information 3: 8e49

Additional Information 4: 62f6201289deb9c567fb9c6d7b0c7c2c

Read our privacy statement:

http://go.microsoft.com/fwlink/?linkid=50163&clcid=0x0409

 

I never played MP before so can't say for sure if this is patch related or not, previously in SP I would typically only crash due to anti-alias (which is off) and/or NVG hangs.

 

As was recently asked, I used to run only firewall / windows defender for protection (seeing as I never got infected anyway, I run a tight network) but did install AVG9 about a month ago for kicks, de-installed it 24 hours ago due to performance impacts and incompatibilities with IE8.

 

The BS MP crashes occurred both before and after I de-installed AVG9.

 

Latest .crash file contained this, but I have about 6-7 others if interested (haven't started to look in them yet even, just let them build up :)).

 

# -------------- 20100804-000727 --------------

C:\Windows\system32\kernel32.dll

# E06D7363 at 75B8FBAE 01:0003EBAE

#

# EAX = 0012F688

# EBX = 727A0CE0

# ECX = 00000003

# EDX = 00000000

# ESI = 7C380EDC

# EDI = 0012F718

# CS:EIP = 001B:75B8FBAE

# SS:ESP = 0023:0012F688 EBP = 0012F6D8

# DS = 0023 ES = 0023 FS = 003B GS = 0000

# Flags = 00000216

 

Ok.. I guess that will get us started, will add more if I remember/find it :)

[topdog]

Virtual store not in use; the game is installed to a location with full current user permissions.

 

No overclocking of cpu/ram/graphics in use. Typically I would measure my system uptime as 'weeks' and would be 'months' except for necessary reboots due to OS patches requiring it.

 

Game is played full screen in a single monitor (though I do have 2 attached to the machine, fullscreen option is still enabled for just primary monitor when playing the game). I've seen others with the protect.dll errors being reported but as far as I recall, not one time has it been that module for me.

Edited August 4, 2010 by topdog

[KeyCat]

Hi topdog,

 

Sounds quite similar to the CTD problems I (and others) experiencing even if I never seen reference to "btmmhook.dll" in my crash-logs, see this thread.

 

Anyway, do you know if your CTD's happens when simulator.exe reaches a specific limit of RAM usage? For me the limit is around ~1.35 GB RAM usage (hard to say exactly since I can't see taskmanager while flying) and then I CTD no matter what. It happens in both MP and SP but RAM usage seems to climb faster in MP (seems dependant on number of players etc).

 

I don't know if your problem is same as the one discussed above but it would be very interesting to know if yours also is related to a specific RAM usage limit of simulator.exe or something else?

 

/KC

Edited August 4, 2010 by KeyCat

[topdog]

32-bit processes can't consume much more than this amount anyway (the 2GB per process is a theoretical maximum, reality is that the amount is usually less) but even if it is a memory issue, it shouldn't (be designed to) just dump out on you like that as an unhandled exception. That's just not how memory management is to be done in the event of a *alloc() being unable to satisfy the request, so would be a bug.

 

I expect they are somewhat related, and I have seen those other threads including yours. In case there are different causes or a solution for one doesn't work for another, each instance needs reporting as a separate thread though really. It was odd in my case that it so frequently would bomb on that module in particular compared to others.

 

I'll look for memory leak type behaviours next time I play though, I appreciate the tip, and it may be something that can be tuned better in my configuration to reduce the load to make more room for the MP fluff in memory so I'll give it a try, thanks.

[Panzertard]

Now thats a new variant, indeed.

From the great interweb:

"btmmhook.dll is a BtMmHook" "belonging to Bluetooth Software" "from ...

So basicly - why would your bluetooth software try to hook onto FC/DCS?

Perhaps you could try to uninstall it to see if you get rid of the problem?

[topdog]

I think I know why; on startup BS iterates over a number of potentially available input devices (even if you don't have one, it looks specifically for trackir, for x52pro, etc.).

 

So it's loaded simply because it's looking for (bluetooth or other) mice/keyboard inputs I guess. The stack does monitor keyboard inputs even from non-bluetooth devices (provides on-screen overlays for when capslock, etc., is pressed), but that shouldn't be directly accessible from BS as it's out-of-process. But just to indicate how it can be relevant to input device drivers that a game would use.

 

I don't really use the bluetooth stack anymore on that machine so I'll uninstall it too, see if that helps or just deflects the failing module to another.

 

Cheers.

Edited August 4, 2010 by topdog

[astrospud]

Can I ask, Topdog, does your game CTD when you are hit or just as you are destroyed?

It has been happening to me lately (5 times, last count). I, too, run Vista 32 and have noticed the only time it does it is when I am taking rounds or direct (missile) hit. I don't acutally see these events, I am 99% sure that's the situation when the CTD happens.

I can only think that damage modelling and particle physics is being modelled, along with other usual flight model things, thus tipping the RAM beyond what I normally experience in level flight or ranged targetting.

Saying this, I have flown in a formation of at least 5 other players over large city (Krasnodar) and although I have a rather low frame rate, the program is still stable.

 

EDIT: Crash and error files attached.

DxDiag.txt

client-20100801-180312.miz.rar

DCS-20100801-091945.crash.rar

network-20100801-1721.log.rar

Edited August 4, 2010 by astrospud
attached error log and crash data

[Panzertard]

Astrospud (hey, funny nick heh)

Missiles & Explosions CTD's can often be linked to effects - and the sound system / loading of graphics.

So if you try to troubleshoot in that direction, check

- Sound system codecs (videoplayers / DVD software often install such) or drivers.

- Also Chipset drivers for the mainboards have been a factor previously.

- and of course, graphics drivers. (most have already tried replacing those).

[astrospud]

Astrospud (hey, funny nick heh)

Missiles & Explosions CTD's can often be linked to effects - and the sound system / loading of graphics.

So if you try to troubleshoot in that direction, check

- Sound system codecs (videoplayers / DVD software often install such) or drivers.

- Also Chipset drivers for the mainboards have been a factor previously.

- and of course, graphics drivers. (most have already tried replacing those).

 

lol! I like it too. Big space/scifi fan and love my potatoes as well...why not combine!

Anyhoo, crash files attached to original post..

Sorry Topdog, don't mean to hijack, but since we seem to have a similar problem...;)

[Panzertard]

astrospud, you PC hardware & such looks good.

I did notice you do use a sounddriver from 2008 (or 2009 depending on which devices you use) - and some bluetooth mic & headset?

Again, combined with some of the effects - perhaps this combination can cause issues.

- Check if Dell or the sound vendor have newer drivers, it's definitively worth checking.

And check for any other software that may interfere with your sound/graphics. If you have videoplayers/editing software and such, these may have installed extra compnents which may interefer. Also worth checking out.

 

Unfortunately the crash-file only stated that Windows didnt cope with the situation - no details from DCS itself.

[Panzertard]

If you guys have CRASH files with other / extra content than just the Kernel32.dll issue may help us a bit more.

 

And for differensial diagnosing:

- Disconnect, disable the extra sound devices - use the builtin instead.

For wireless / BT equipment - be sure the software isnt loaded (device disabled + killing the startup applications for it).

 

At least it may provide us information if it changed anything or not.

Note, some people are not able to start DCS without any sounddevices at all - so you may to have at least one sounddevice active.

Edited August 4, 2010 by Panzertard

[topdog]

I've been flying online in formation with other choppers at the time I crashed, so I wasn't shot down or under attack, it would have been seen and they would have been next.

 

I haven't been able to identify yet anything common about the occurrences; they're all inside the chopper (since external views are disabled), they're during daytime conditions (no ghost of NVG/AA pasts), and the only graphics anomaly that seems new to me but I think has been dsicussed elsewhere on the forums is the issues with the water and coastline - it looks like the land is floating above the water surface and there's no coast effects, but I have crashed when no water is visible (even though we all know the water runs right through and under the earth).

 

So no luck on a commonality there, but I haven't yet done the memory monitoring or disabling of BT stack (I'm still at work currently) so I'll still be able to give those a shot.

 

Sound in my system is pretty good (compatibility wise, technically it's just a soundblaster x-fi gamer, basic, but is robust and does the job), and the only sound issues I ran into somewhat recently were resolved with c0ff's help for FC2, which was more a case of either having sound or none at all, but no crashes. That and the mysterious engine 2 vs. engine 1 startup issue, which isn't really significant and everyone gets that anyway.

[topdog]

Ok, interesting development after uninstalling the bluetooth stack. What I was getting before (consistently) were 2 concurrent crashes (and windows dialogs to match them) when my game ended.

 

One was always the btmmhook.dll, and the other was just 'some random DCS module' (edcore, mitkagraphics, etc.). Now I just get the random module one when it crashes alone now.

 

 

So, a worthy test to do, but looks like a it was just a victim rather than the cause this time around.

 

I also confirmed the complete removal of the bluetooth stack drivers from my system, btmmhook.dll which was in my %windir%\system32 was gone after the uninstall and reboot that it requested.

 

I didn't pay attention to memory at this time, I was just content to fly after the uninstall without paying too close attention to things, to see if I would encounter the problem. It was again after about 2-3 hours of flying, which fits into my normal yet somewhat random timeframes to get it.

 

After the crash I popped into my temp directory and did "dir /od" to locate the newest files, the reason for the time lag between the .crash file and the others is simply because I was still talking to 2 other pilots in my flight group to let them know what's up :( C'est la vie.

 

...
04/08/2010  08:20 PM    <DIR>          ..
04/08/2010  08:20 PM    <DIR>          .
04/08/2010  08:20 PM               398 DCS-20100804-192038.crash
04/08/2010  08:23 PM           162,967 Error.log
04/08/2010  08:23 PM           767,823 debrief.log
04/08/2010  08:23 PM           146,001 obj_arrays.log
04/08/2010  08:23 PM        76,390,400 ~tr48AB.tmp
04/08/2010  08:23 PM       219,500,544 ~tr48AA.tmp
            274 File(s)  1,394,074,091 bytes
              3 Dir(s)   5,916,504,064 bytes free

C:\Program Files\Eagle Dynamics\Ka-50\Temp>type DCS-20100804-192038.crash
# -------------- 20100804-192038 --------------
C:\Program Files\Eagle Dynamics\Ka-50\bin\x86\stable\edCore.dll
# C0000005 ACCESS_VIOLATION at 0027169F 01:0001069F
#
# EAX = 00000000
# EBX = 03C9AE24
# ECX = 00000000
# EDX = 101E0B08
# ESI = 0027D380
# EDI = 3D62241C
# CS:EIP = 001B:0027169F
# SS:ESP = 0023:3E75F26C  EBP = FFFFFFF2
# DS = 0023  ES = 0023  FS = 003B  GS = 0000
# Flags = 00010246

C:\Program Files\Eagle Dynamics\Ka-50\Temp>

 

From the Problem Reports applet (more or less the same thing):

 

Product

DCS Black Shark

Problem

Stopped working

Date

04/08/2010 08:20 PM

Status

Not Reported

Problem signature

Problem Event Name: APPCRASH

Application Name: simulator.exe

Application Version: 1.2.1.0

Application Timestamp: 4bed01e6

Fault Module Name: edCore.dll

Fault Module Version: 0.0.0.0

Fault Module Timestamp: 4becff7a

Exception Code: c0000005

Exception Offset: 0001169f

OS Version: 6.0.6002.2.2.0.256.1

Locale ID: 2057

Additional Information 1: a34a

Additional Information 2: c9c5f4fd744690d388ab9d5b3eb051a7

Additional Information 3: bf4a

Additional Information 4: f1ea91ae32511d63575c3b59717fed46

　

 

 

I was just moseying down towards a waypoint at a reasonable sub-140k's cruise, nothing engaged me or shot me down (one of my flight group was slightly behind me and had visual).

 

Edit: P.s. Program Files / Virtualstore isn't the problem..... :)

Edited August 4, 2010 by topdog

[Panzertard]

The coast / water is like that, "it floats".

 

Now you seem savvy enough to have poked around a bit already.

- How much physical memory do you have in total?

- Care to dump us a DXDiag report?

- Have you tested with a vanilla reinstall - No mods installed afterwards?

You can rename your old Ka-50 dir to "ka-50 backup", then install a new one in the location of the old - an patch it up to 1.0.2. That way we would know it's no mods / other config contributing to the issues.

[astrospud]

astrospud, you PC hardware & such looks good.

I did notice you do use a sounddriver from 2008 (or 2009 depending on which devices you use) - and some bluetooth mic & headset?

Again, combined with some of the effects - perhaps this combination can cause issues.

- Check if Dell or the sound vendor have newer drivers, it's definitively worth checking.

And check for any other software that may interfere with your sound/graphics. If you have videoplayers/editing software and such, these may have installed extra compnents which may interefer. Also worth checking out.

 

Unfortunately the crash-file only stated that Windows didnt cope with the situation - no details from DCS itself.

 

Cheers Panzer. Gives me a good idea where to go. Many thanks! :)

[topdog]

The coast / water is like that, "it floats".

 

Now you seem savvy enough to have poked around a bit already.

- How much physical memory do you have in total?

- Care to dump us a DXDiag report?

- Have you tested with a vanilla reinstall - No mods installed afterwards?

You can rename your old Ka-50 dir to "ka-50 backup", then install a new one in the location of the old - an patch it up to 1.0.2. That way we would know it's no mods / other config contributing to the issues.

 

4GB physical, but as you'll see in the dxdiag it equates to 2.8GB available for OS and apps (32-bit OS, subtract 512 for OS hardware address range mapping, and subtract 768 for graphics RAM address range mapping). Sucks to lose that much I know, but at the time Vista 64 was untested and drivers were especially bad, so it was common sense to stick with 32-bit even with that much RAM.

 

The other tests (reinstalling, for which I'll need to make some room first, and monitoring memory usage for leaks) will take some time so might be a weekend job before I get back to the thread on those.

 

I will remind you though that tacview only puts 1 line into export.lua and then references it's own .lua for everything else, and in my own hand-written modding I followed the same style, so uninstalling mods was fairly simple and complete to do. I haven't screwed around with additional models, skins, or changing of game file parameters, with one other exception and that was adding my own 'radio maykop' beacon up the top of a mountain. I'll check to make sure that's gone too but I believe it was overwritten in the patch anyway, and it'll be extra moot when I do the reinstall tests this weekend.

 

By the way; if I move my current install and try to reinstall in-place to my main location, do you happen to know how that might behave with respect to activations, and should I do anything else first like deactivate or not need to bother with that? It's not a big concern anyway, just curious to know what I should expect.

 

Thanks.

DxDiag.txt

Edited August 5, 2010 by topdog

[Panzertard]

No, you shall not lose any activations for reinstallations - the key is stored in registry. It may ask for activation, but will discovered that it is already registered.

 

Yes, try not to use Tacview in your limited memory environment - it may improve your situation.

Looking forward to hear some results :)

[astrospud]

I hope you don't mind me piggybacking your thread, TD. It seems our probs may be similar, but all things being as they are, our machines are different.

 

I followed Panzers' advice and shut down any blue tooth devices I came across. I attempted to update the sound device driver, but the latest is from 2008...

 

A note on my install, I do run the current Modman with a few mods installed:

 

 

Could these be an issue?

 

From Windows Error reporting:

Product

DCS Black Shark

 

Problem

Stopped working

 

Date

8/08/2010 7:48 PM

 

Status

Not Reported

 

Problem signature

Problem Event Name: APPCRASH

Application Name: simulator.exe

Application Version: 1.2.1.0

Application Timestamp: 4bed01e6

Fault Module Name: protect.dll

Fault Module Version: 5.70.13.2

Fault Module Timestamp: 4bed03ee

Exception Code: 40000015

Exception Offset: 00e4ab94

OS Version: 6.0.6002.2.2.0.256.1

Locale ID: 3081

Additional Information 1: 696e

Additional Information 2: b69fbaac43282bb7fb6a4cb572aac8b7

Additional Information 3: af39

Additional Information 4: 0d8b83a94eb08861c5fbaa7229cea18d

 

 

I've attached the crash files.

 

Cheers

DCS-20100808-101820.rar

Error.rar

errors.rar

DxDiag08082010.txt

[Panzertard]

astrospud, I cannot see any new solutions with your problems.

And yes, mods can very much contribute to crashes - and its so hard to determine which of them, or which combination.

As well as some of the mods / modman may have left content that you cannot fix again later.

 

So if you too could run a vanilla verion, no mods - that way you can eliminate if the mods contribute to your problem or not.

[Hower]

Re: Crashing BS in MP mode

 

As i advice you yesterday night on 104th server for other people hanging with similar crashes in multiplayer mode. My solving is to switch the game in options to not using fullscreen mode. !!! Turn off fullscreen mode in options !!!

 

It helped me to overcome many problems from graphical artifacts to overheating the graphic. I hope that will solve this thread anyway is a juvenile solution but gogo...

 

"you have to set the same resolution and screen depth as your native desktop resolution/screendepth as you desktop monitor and you will notice no diference between full screen and windowed mode"

 

howgh:book:

[topdog]

I can give that a go easily enough, but I'm experiencing no overheating or graphic artifacts during play (or crash). Thanks Hower.

 

I just had another crash (and another last night, 2 different maps, and not many players on right now, about 12-14), and had plenty of memory headroom at the start and at the point of crash, didn't appear to leak.

 

Here's a summary so far.

 

- OS is patched up

- over 15GB free space

- last defragged on wednesday (scheduled weekly)

- no mods

- no antivirus / crash guard type software currently, uninstalled avg9 about a week ago (was crashing before and after this and wasn't the reason I uninstalled it)

- no virtualstore/uac file permission issues

- fc2 seemingly runs fine without crashing (though I don't do multiplayer in fc2)

- bs 1.0.1c ran fine without crashing (exception: 2 missions known to crash, especially w/nvidia cards, such as night-ops mission with nvg and AA enabled, though again this wasn't multiplayer)

- crashes with AA forced on or off

- sufficient available memory headroom

- not close to hitting 2GB max per process

- not due to MP map being played for too long

- dxdiag provided

- doesn't crash during other games or intensive applications (e.g. dbms, app compiling, 2009 release games)

- doesn't matter if DEP is enabled or disabled (either system-wide, or just on simulator.exe)

- system heat stable, not overclocked anywhere (not on ram, cpu, graphics)

- 195.62 and 258.96 nvidia drivers both tried

- bluetooth stack uninstalled upon request

- game reinstalled from request (english dvd + md5 checked 1.0.2 patch english patch file)

- had to remove HKLM\SOFTWARE\Eagle Dynamics\BlackShark\PatchVersion before it would install because it said I had 1.0.2 already

 

 

 

 

 

 

 

 

 

 

Most recent:

 

 

The second you see the CPU usage nosedive is the point in time of the crash occurring.

 

 

C:\Program Files\Eagle Dynamics\Ka-50\Temp>type DCS-20100808-144642.crash

# -------------- 20100808-144642 --------------

C:\Windows\system32\kernel32.dll

# E06D7363 at 7763FBAE 01:0003EBAE

#

# EAX = 0012F620

# EBX = 75952AA0

# ECX = 00000003

# EDX = 00000000

# ESI = 7C380EDC

# EDI = 0012F6B0

# CS:EIP = 001B:7763FBAE

# SS:ESP = 0023:0012F620 EBP = 0012F670

# DS = 0023 ES = 0023 FS = 003B GS = 0000

# Flags = 00000212

 

 

E06D7363, I am lead to understand, is an exception message from the MSVC runtime when a 'throw' occurs to give the app/developers a chance to deal with it. The latter 3 bytes '6D7363' are 'msc' in ascii, which is a code they use to signal this.

 

Next and final test available, not using fullscreen mode.

Edited August 8, 2010 by topdog
Putting most recent info in correct place of post chronologically

[topdog]

No luck with windowed mode.

 

Practically identical experience as before, only difference being a loss of framerate whilst playing (which was to be expected).

 

C:\Program Files\Eagle Dynamics\Ka-50\Temp>type DCS-20100808-171501.crash
# -------------- 20100808-171501 --------------
C:\Windows\system32\kernel32.dll
# E06D7363  at 7763FBAE 01:0003EBAE
#
# EAX = 0012F688
# EBX = 3F26FA28
# ECX = 00000003
# EDX = 00000000
# ESI = 7C380EDC
# EDI = 0012F718
# CS:EIP = 001B:7763FBAE
# SS:ESP = 0023:0012F688  EBP = 0012F6D8
# DS = 0023  ES = 0023  FS = 003B  GS = 0000
# Flags = 00000216

C:\Program Files\Eagle Dynamics\Ka-50\Temp>

 

Product

DCS Black Shark

Problem

Stopped working

Date

08/08/2010 06:15 PM

Status

Not Reported

Problem signature

Problem Event Name: APPCRASH

Application Name: simulator.exe

Application Version: 1.2.1.0

Application Timestamp: 4bed01e6

Fault Module Name: edCore.dll

Fault Module Version: 0.0.0.0

Fault Module Timestamp: 4becff7a

Exception Code: c0000005

Exception Offset: 0000685b

OS Version: 6.0.6002.2.2.0.256.1

Locale ID: 2057

Additional Information 1: 7c21

Additional Information 2: ffd7e05752ea605490b36a36833228a8

Additional Information 3: eb6c

Additional Information 4: 7b34ee80b4be82fdb698dff618a23957

Further tests I intend to do (try and build a fuller picture and/or narrow scope of problems), but otherwise at a loss for the moment:

 

- go back to some single-player playing and see what it's like

- do some multiplayer fc2 on same server and see what it's like

Edited August 8, 2010 by topdog

[topdog]

So far, no crashing in single player mode (was just doing the Quick Mission and flying around for an hour+).

 

I did crash it once in local server multiplayer mode, on the same map that I crash on on 104th as a few copies got left behind in my temp folder after the crashes, and ended up with a similar but not quite identical .crash dump file.

 

The odd thing about these tests is that my memory usage with this mission on 104th is ~1250MB to 1450MB. When playing the same map in multiplayer locally it is consuming ~1350 to 1550MB. When playing the Quick Mission it's ~1500 to 1600MB used.

 

So it's not even memory accumulation or a 'leak' of memory not being deallocated causing it.

 

Still testing I guess, as it takes time to build up the results, but wishing there was something better for debugging built in I could enable, to capture what's needed to identify the cause, as it's a royal pain.

 

My online experience is akin to spending 30 mins flying farp to airfield, get shot down, flying farp to airfield, crash to desktop, then flying farp to airfield again... one hour 30 mins later, I get to fire my first shots, the team on the other side have already claimed their airport and are cleaning up the WP groups, and I'm about to start throwing heavy objects around the room :)

[Hower]

Memory accumulation

 

TopDog> sry for later reply but i trying to do my best.

 

So it's not even memory accumulation or a 'leak' of memory not being deallocated causing it.

 

From screens i asume u are running an Win7 system with i`m not so familiar but your ram drain is bit high no? I now that system is cappable of running more RAM than XP Pro 32 bit/SP3 but i calculate complete system resources not only gaming consumption. I`m not an IT Expert but my clean Win start after boot has 275MB of PF after TS3 Client and FreeTrack load it loads another 50-100MB and when playing 104th servers i get sometimes crash too but first i notice dramatic framerate los. I started above 30FPS after an hour or two their simply drop belov 10FPS unplayable. I asked on chat what to do while the game was nearing to crash and someone answered me to restart the game completely and cleaning the pagefile. That have to do with realtime performance. I loose all credits but the game start as new over 30FPS. Ya i now it is low but i running it from Alienware laptop

 

On 104th servers over 16 players i never going over 1.5 GB of ram usage for complete system the simulator.exe alone takes never more than 1.2GB. I noticed when the pagefile is larger than my ram amount, the system is trying to write the data left on the hardisk. U can tweak it...

 

But such demanding GIS data in this game is maybe overstressing the disks. I upgraded the ram to the max of XP Pro Win32 bit to 2+2GB and no problems occurs more. Only thing left is writing the *.trk replay tracks because their size is somethimes very big. I clean them manualy after playing from "BlackShark/Temp" folder

[topdog]

Hi Hower, have been a bit busy myself lately and these forums are a timesink (i.e. very active, many posts/threads,) so I had to curfew myself :)

 

I'm using Vista 32-bit, but my physical memory isn't being saturated here so paging and/or being out of physical memory isn't causing it. The overheads of newer OS versions than XP are indeed higher though, and that is normal / to be expected.

 

As I've shown, I can make the game use more memory and still run fine (1500-1600MB) than what it is trying to use at the times it crashes - it is hundreds of megabytes difference, or hundreds of MB left of room to play in, between the game sessions and yet it can still CTD.

 

Tez asked me if I could provide a track file of the crashing (and to see if from that, a replayable scenario causing the crash at the same place and time can be created), but I haven't been able to expend much energy on doing this.

 

For one thing, when it crashes to desktop, you don't get a track file - it cannot compile (zip) the necessary files into a complete .trk file, and my attempts to cobble one together with the remnants of the /temp directory have not been successful either.