Jump to content

(MT) Game crashes in the first minutes of joining a MP server due to a memory leak


Recommended Posts

Posted (edited)

Had this problem since the first MT version, happens in VR and it sends my headset to sleep, kills windows explorer and i have to restart my PC to bring it back. Event viewer says: ''Windows successfully diagnosed a low virtual memory condition. The following programs consumed the most virtual memory: DCS.exe (19120) consumed 38467153920 bytes, dwm.exe (1532) consumed 1459068928 bytes, and ST.exe (3412) consumed 705662976 bytes.'' I have 16GB of DDR4 ram and 32GB pagefile on an SSD. Can't play on ST either since the first MT update, i'm getting a weird fishbowl distortion in my headset. While it works I get 50ish FPS with my Reverb G2 and 3080 which is amazing while it lasts

Do let me know if and which files u need me to upload

Btw I did a full repair and cleanup, as well as a reinstall when the repair failed. Only thing left i can do is windows reinstall which i dont want to do since i use this PC for work as well

dcs.log dcs.log-20230423-173450.zip

Edited by bukizzzz
  • bukizzzz changed the title to (MT) Game crashes in the first minutes of joining a MP server due to a memory leak
Posted

It looks like a memory issue, indeed.

You should consider upgrading to 64GB of RAM, or at least 32GB.

In the meantime, try lowering DCS preload radius.

2023-04-23 17:34:26.281 INFO    ASYNCNET (Main): Current ping: 48.7ms
2023-04-23 17:34:33.194 WARNING WORLD (Main): ModelTimeQuantizer: SAME MODEL TIME
2023-04-23 17:34:36.634 ALERT   EDCORE (19652): Out of memory in ed_malloc for 22369828 bytes.
2023-04-23 17:34:36.650 INFO    EDCORE (19652): try to write dump information
2023-04-23 17:34:36.766 INFO    EDCORE (19652): # -------------- 20230423-173436 --------------
2023-04-23 17:34:36.767 INFO    EDCORE (19652): DCS/2.8.4.39313 (x86_64; MT; Windows NT 10.0.22000)
2023-04-23 17:34:36.767 INFO    EDCORE (19652): C:\Program Files\DCS World OpenBeta\bin-mt\edCore.dll
2023-04-23 17:34:36.768 INFO    EDCORE (19652): # 80000003 BREAKPOINT at 00007ffef38676bb 00:00000000
2023-04-23 17:34:36.771 INFO    EDCORE (19652): SymInit: Symbol-SearchPath: 'C:\Program Files\DCS World OpenBeta\bin-mt;', symOptions: 532, UserName: 'Vuk'
2023-04-23 17:34:36.772 INFO    EDCORE (19652): OS-Version: 10.0.22000 () 0x100-0x1
2023-04-23 17:34:36.778 INFO    EDCORE (19652): 0x00000000000276bb (edCore): ed::fatal_out_of_memory + 0x3B
2023-04-23 17:34:36.779 INFO    EDCORE (19652): 0x0000000000022c77 (edCore): ed::SharedBuffer::create + 0x37
2023-04-23 17:34:36.780 INFO    EDCORE (19652): 0x000000000005e658 (edCore): edvfs::FSMappedFile::tell + 0x10A8
2023-04-23 17:34:36.784 INFO    EDCORE (19652): 0x000000000005e42c (edCore): edvfs::FSMappedFile::tell + 0xE7C
2023-04-23 17:34:36.784 INFO    EDCORE (19652): 0x0000000000055533 (edCore): CoreUtils::TempFilesManager::TempFilesManager + 0x3BB3
2023-04-23 17:34:36.785 INFO    EDCORE (19652): 0x000000000000752a (renderer): render::openFirstOccurrence + 0x8A
2023-04-23 17:34:36.786 INFO    EDCORE (19652): 0x0000000000062e57 (dx11backend): createRenderer + 0x4AF47
2023-04-23 17:34:36.786 INFO    EDCORE (19652): 0x00000000000662eb (dx11backend): createRenderer + 0x4E3DB
2023-04-23 17:34:36.786 INFO    EDCORE (19652): 0x0000000000036305 (dx11backend): createRenderer + 0x1E3F5
2023-04-23 17:34:36.786 INFO    EDCORE (19652): 0x0000000000033c39 (edCore): ed::this_thread::yield + 0xB59
2023-04-23 17:34:36.787 INFO    EDCORE (19652): 0x0000000000049228 (edCore): ed::this_thread::yield + 0x16148
2023-04-23 17:34:36.787 INFO    EDCORE (19652): 0x00000000000326b1 (edCore): ed::thread::_get_current_thread_id + 0x71
2023-04-23 17:34:36.788 INFO    EDCORE (19652): 0x0000000000026c0c (ucrtbase): _recalloc + 0x5C
2023-04-23 17:34:36.789 INFO    EDCORE (19652): 0x0000000000015590 (KERNEL32): BaseThreadInitThunk + 0x10
2023-04-23 17:34:37.624 INFO    EDCORE (19652): Minidump created.

 

  • Thanks 1

---

Posted
On 4/23/2023 at 2:09 PM, Flappie said:

It looks like a memory issue, indeed.

You should consider upgrading to 64GB of RAM, or at least 32GB.

In the meantime, try lowering DCS preload radius.

 

Well, if it is indeed a memory leak, won't upgrading to 64GB just delay the crash a little longer?!

Posted (edited)

If that is indeed a memory leak issue, it will get fixed at some point, of course.

But even without any memory leak, DCS takes a lot of RAM. My PC has 32GB of RAM, still it's not enough when playing over Syria in MP, for instance. This is were the pagefile is useful, but paging has an impact on performance (RAM will always be faster), which is why I suggest to go for 64 GB.

Edited by Flappie

---

Posted
On 4/28/2023 at 3:01 PM, Flappie said:

If that is indeed a memory leak issue, it will get fixed at some point, of course.

But even without any memory leak, DCS takes a lot of RAM. My PC has 32GB of RAM, still it's not enough when playing over Syria in MP, for instance. This is were the pagefile is useful, but paging has an impact on performance (RAM will always be faster), which is why I suggest to go for 64 GB.

 

I'd better start looking at my logs and see what the heck is going on occasions.  I only use ST DCS.exe, single play, loaded with 64g DDR5 (fixed page file set to 9216, never even bothered above that).  I haven't had any crashes in years, even before my recent rig upgrades, until this MT thing showed up. I gave up on it for now. Memory was probably not the only thing leaking there, hehe.

Currently ST is holding up OK for me, even in fairly long missions but... once in a while my Reverb G2 goes 'grey'. Nothing will fix it until the OS reboot. The game is still running on the pancake screen and the head tracking still works.  There is some other strange behavior after quitting the game or rebooting my rig but I have to investigate more, could be just VR related.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...