Samsung Magician 4.5 memory leak, caused by RAPID-Mode & possibly KB3004394 (Windows 7/8/8.1 update)

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630
Hey,

I have no idea about how RAM is supposed to handle when you reach your RAM capacity, but whenever my computer reaches it (say, 15.6GB of 16GB) apps will start crashing out of the blue.

Not only this, but then eventually something big like chrome crashes, and then suddenly my entire computer turns black, as if the nvidia drivers crashed or something.


Here are my specs:

ASUS Maximus VI HERO (supports up to ddr3-1600, anything higher is reported as an overclock)
XFX Pro 1000W LTD Black Edition Power Supply
i5 4790k haswell CPU
4x4GB Ripjaws X DDR3-1600 memory (no OC)
Samsung SSD 840 Pro Series (250GB, with Windows 8.1 installed on this SSD)
Samsung SSD 840 EVO (750GB)
1TB HDD from my old build
GTX 780 (non-reference card, second best possible OC on market I believe)


Does anyone have an idea if this is normal for when you're nearing the max capacity for your RAM? Or is this not normal? Please help!
 
1. Your Hero supports up to DDR3-3000 ... When you buy RAM, the speed written on the package is generally the XMP profile which a) is the rated speed of the RAM, and b) is *officially* an overclock cause there is no JEDEC sanctioned profile for it. The Hero is perfectly suited to run any of the speeds below though I'd have a hard time recommending over 2400.
http://www.newegg.com/Product/Product.aspx?Item=N82E16813131989
DDR3 3000(O.C.)/2933(O.C.)/2800(O.C.)/2666(O.C.)/2600(O.C.)/2500(O.C.)/2400(O.C.)/2200(O.C.)/2133(O.C.)/2000(O.C.)/1866(O.C.)/1800(O.C.)/1600/1333

2. How did the 4 x 4 Gb come about ? .... one kit or two ? .... if you married two separate 2 x 4Gb kits, that would be my best guess as to the source of the problem. Using 4 sticks often requires increased DRAM voltage, relaxed timings and / or speed.

3. If the sticks are crashing, run Memtest86+ at least overnight with all sticks in place.

4. How are you managing to "use" 15.6 GB .... RAM Disk ?

5. When you say "2nd best OC on the market", are we talking what it it says on the side of the box (meaningless) or a manual OC ? The better 780 should have no problem hitting 1250 boost clock (1075 - 1125 clock).

6. If you load all your programs that normally cause this condition , does the RAM usage stay the same or increase the longer the PC is on (memory leak) ? How often do you reboot machine ?

7. Where are Page and temp files ? 750 GB SSD I presume.

 

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630


1.) and 2.) Oh, okay. I can definitely try getting a 2x8GB DDR3-2400 then, later, if it is indeed memory causing issues.

3.) Actually, the sticks don't "crash." Just windows apps. Never had a BSOD.

4.) Heh. Well, I have the same question, actually. I will have to check my RAM every once in a while. Right now it is at 9/10GB usage, which to me seems quite normal. 15.6GB *is* really odd.

5.) This is my video card.

6.) We'll have to see. I'll make it a point to have task manager open to observe if or when my memory goes above 10GB, to see if there is a memory leakage as the culprit.

7.) Nope, my steam games are on the 750 GB SSD. The OS is on the 250GB one (113GB free on it.)



I was nearing my RAM's maximum capacity, and it did not BSOD. Rather, it would cause programs to suddenly stop responding and have to be closed. Realtemp is one that often crashes first. Then, eventually, I will get the "Oops! Chrome has crashed..." message, and then I will have to restart Chrome. (No idea if chrome.exe is the one causing the memory spike, but of course I will have to check) and eventually, something will crash and all four of my monitors will turn black. The computer hasn't crashed at this point (no BSOD)... just some graphics glitch, but what ends up happening as well is my mouse cursor disappears and I cannot ctrl + alt + del or alt + tab or anything. Likely at this point some important system driver crashed and I have to restart my computer.
 

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630
Beware... plot twist incoming!!


It happened again. Memory usage was ~11GB.

First, my alt-tab stopped working for some reason. I could tab between the previous visited window, but pressing tab multiple times did not work. Wish I had taken a video snapshot of it, as it was really odd. Then, to no avail, I started a game (uses mainly opengl stuff) and then poof! My Python editor program crashes out of nowhere.

But that isn't where it stops. I close the game, and re-open it again, and then all of my monitors turn black, and the only thing that's left is a blinking mouse cursor. I could actually *move* the cursor. Not only that, but my Pandora music was still running in the background! So, my guess is that everything else was still running perfectly, except I cannot navigate anywhere.

The alt-tab breaking I think may be a huge clue. Could it be an OS related problem? Also, I think I have to be extra careful here. Just because my computer became completely inaccessible when I opened up an OpenGL program doesn't mean that it's a GPU related issue. Perhaps the only reason it happened was because it happened to be running at the wrong place at the wrong time, and it became victim to this weird process-crashing issue, except instead of any random process, it's some essential one that, without it, causes your whole computer to turn black and all that's left is a useless blinking mouse cursor.


I will run memtest before I go to bed.
 

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630
Hi,

I ran memtest86+ overnight. Ran for 5 hours 44 minutes, 4 passes on all 10 or 11 tests and got no errors.

I am thinking this is a driver issue. Think about this: what driver would cause the alt+tab to stop functioning properly? That happens every single time.

The strange thing is, right when I reboot my machine, alt-tab works perfectly. *Something happens* while it's on that causes it to break.

I decided to play around with my computer before the test and I logged into another account while it presented the most symptoms. It didn't even load properly. The computer BSOD'ed with message SYSTEM_SERVICE_EXCEPTION (faulting driver, ntoskrnl.exe.)

I need as many possible ideas as possible. Look above for my full computer specs.

Thanks,
dazinger
 

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630
Seems yesterday, I got a LOT of this error (it was spammed perhaps 30+ times in the course of a few seconds, at 11:29:13 AM-11:29:15 AM yesterday, and also it happened apparently on the 23, 24, and the 25th.


The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

\Device\Video3
Graphics Exception: ESR 0x404490=0x80000002

Some display the last line as Graphics Exception: EXTRA_MACRO_DATA or Graphics Exception: Shader Program Header 11 Error... doesn't seem to be any other different variations.

Eventviewer views this error 335 times over the past 7 days. Not sure what happens at these specific times, but maybe it causes that black screen / flickering cursor bug.. but this seems like a big clue.



The application crashes that I described above also show up in the Event Logs. Faulting module path is either the application itself or KERNELBASE.dll.
 

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630
Regarding NVIDIA, I did go about installing 347.09 drivers today. Actually, what I did was I checked to see maybe if my last driver installation was broken somehow, so I went back and installed the NVidia 340.82 driver from the OpenGL Driver page on NVidia's site to see if that would work. It installed successfully, then went to Geforce Experience to update it back to 347.09.

So far everything looks normal. Alt tab working as expected.


Though, could you try to figure out what that above nvidia error could mean?
 

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630
Well.... the computer seemed perfectly fine since my last post, and it was fine all of yesterday, but just 15 minutes ago the system crashed. Not a BSOD, but it may as well have been one.

One clue I got was, Skype was apparently saying it had been terminated "due to low system resources." RAM usage was <10GB usage, and CPU seemed fine.

I'm not sure how to track video card usage. Or maybe it has to do with page file? I don't know...

Chrome tends to crash a lot. I just enabled crash reporting, so I should be able to see them the next time it happens.

Checking event viewer again.... the Skype error that I mentioned above is this: "Package Microsoft.SkypeApp_3.1.0.1007_x86__kzf8qxf38zg5c+App was terminated due to low system resources." Happens a lot apparently...


Oh.... Malwarebytes service also crashed a lot. Here is an actual crash log:

Fault bucket 73702023866, type 1
Event Name: APPCRASH
Response: Not available
Cab Id: 0

Problem signature:
P1: mbamservice.exe
P2: 3.0.8.1
P3: 546e4a58
P4: mbamservice.exe
P5: 3.0.8.1
P6: 546e4a58
P7: 40000015
P8: 0008f796
P9:
P10:

Attached files:
C:\Windows\Temp\WER34FC.tmp.appcompat.txt
C:\Windows\Temp\WER354B.tmp.WERInternalMetadata.xml
C:\ProgramData\Microsoft\Windows\WER\ReportQueue\AppCrash_mbamservice.exe_62ae4952fc5873b9c7be325f68291b1757d61dd_608eb1e2_cab_3cf0354a\memory.hdmp
C:\ProgramData\Microsoft\Windows\WER\ReportQueue\AppCrash_mbamservice.exe_62ae4952fc5873b9c7be325f68291b1757d61dd_608eb1e2_cab_3cf0354a\triagedump.dmp
WERGenerationLog.txt

These files may be available here:
C:\ProgramData\Microsoft\Windows\WER\ReportArchive\AppCrash_mbamservice.exe_62ae4952fc5873b9c7be325f68291b1757d61dd_608eb1e2_bfe05f57

Analysis symbol:
Rechecking for solution: 0
Report Id: f003888f-9073-11e4-bf94-e03f4917145a
Report Status: 0
Hashed bucket: 34ea8ce8e7e490884c5c2da0a089d24a
 

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630
Well, I haven't had any random crashes so far today, yet I am getting that Skype "low system resources" error all day.

Still keeping an eye on Chrome -- the first time it crashes, I'll know to look at the crash log as soon as possible.

Event viewer noted a failure last night to update Windows. So, I'll see if that may help... if anything abnormal pops up before then I'll be sure to post it here as an update.

Happy New Year!
 

dazinger92

Distinguished
Nov 20, 2013
57
1
18,630
Chrome crashed... and I found yet another clue..


The NVIDIA OpenGL driver has encountered
an out of memory error. This application might
behave inconsistently and fail.
(stop logging every single event of this type
because there are too many)


This showed up as a warning, surprisingly, so I didn't catch it right away. But it probably explains the graphic glitches.

But still.... hmm....

I'm going to try upping the page file memory to see if that helps. I'm really out of solutions... windows update worked perfectly, with 16 gigs of memory I can't possibly see how it went from 10GB to 16GB but maybe it did and I'm just not seeing it.



Update: I can already tell the solution's probably not going to work. Checked GPU-Z, GPU settings all look normal, CPU normal, memory looks fine, half used up.

This really does stump me. Resource monitor is showing that commit charge is nearing 100% (right now 80%) and my computer seems to be slowing down as I type. The best symptom is the alt-tab function... holding down alt-tab for a bit, and letting go of just tab, you can see that the computer is still performing tab actions... and this does NOT happen at startup.

Now looking at this thread to see if maybe I can somehow track the culprit.

I'm not giving up yet! :)


UPDATE 2: Windows shows commit charge rising over 80%. Standby memory 6574MB and rising, Free memory 899 MB and decreasing... out of 16GB total and 8.6GB of physical memory used, it's about the closest thing I have to figuring out what makes the slowdown increase.

Just installed WDK 8.1... going to run poolmon.exe now.

Code:
NDbf Nonp       1025 (   0)         0 (   0)     1025  1098817536 (          0)
CM31 Paged     67693 (1061)     23356 (   0)    44337   209825792 (    5021696)
NDft Nonp       9808 (   0)      7292 (   1)     2516   187966192 (        -48)
MmSt Paged    262141 ( 210)    218585 ( 208)    43556    69036400 (      10768)
Proc Nonp      25198 (  32)        13 (   0)    25185    48758096 (      61952)
FMfn Paged   1778914 (1187)   1679037 (1171)    99877    43533136 (       6720)
Toke Paged    358946 ( 228)    331533 ( 195)    27413    40119936 (      46240)
Ntff Paged     40751 (   2)     11938 (   0)    28813    37341648 (       2592)

Code:
C:\WINDOWS\system32>findstr /s NDbf *.sys
drivers\SamsungRapidDiskFltr.sys:╛☺  └Θ¬          D9│|☻  uφHïâ░   HïH@ïâp¶  9AH☼
D9│╚↕  ☼äp☺  3╥Aï─3╔≈│î¶  A╕NDbfAì<ï≈Hì¶⌂H┴Γ♥ §iE☻ Hï↔*í☻ Hëâx¶  I;╞u
drivers\SamsungRapidDiskFltr.sys:╕Ü  └Θ ☺  ïüp¶   ╚à╟t☼ïëp¶  ìA #╟+╚♥∙A╕NDbfï╫3╔
Dïτ §°╙☺ Hï≡Hà└t└Hï↔▒/☻ Mï─3╥Hìï°↔  Hï)HïE Hë☺Hë ▲  Hï╬Φ        ░☺ Hâe( ë}►ë}¶Dë
m↑Hëu 3╥KìîmA☺  H┴ß♦H♥╦HïHëM HHë(ï╟Hë≈│î¶  3╥à└t=ï╨Dï└Hï¶  Hà╔t
DRVSTORE\SamsungRap_843B5A6351DBB5C2DE5E15ABB31AA472E0095A78\amd64\SamsungRapidD
D9│╚↕  ☼äp☺  3╥Aï─3╔≈│î¶  A╕NDbfAì<ï≈Hì¶⌂H┴Γ♥ §iE☻ Hï↔*í☻ Hëâx¶  I;╞u↕  u
DRVSTORE\SamsungRap_843B5A6351DBB5C2DE5E15ABB31AA472E0095A78\amd64\SamsungRapidD
iskFltr.sys:╕Ü  └Θ ☺  ïüp¶   ╚à╟t☼ïëp¶  ìA #╟+╚♥∙A╕NDbfï╫3╔Dïτ §°╙☺ Hï≡Hà└t└Hï↔▒
/☻ Mï─3╥Hìï°↔  Hï)HïE Hë☺Hë ▲  Hï╬Φ     ░☺ Hâe( ë}►ë}¶Dëm↑Hëu 3╥KìîmA☺  H┴ß♦H♥╦H
ïHëM HHë(ï╟Hë≈│î¶  3╥à└t=ï╨Dï└Hï¶  Hà╔t
DRVSTORE\SamsungRap_91B2624CEE8F4E65E61BE725467EB7EF656E22D2\amd64\SamsungRapidD
D9│╚↕  ☼äp☺  3╥Aï─3╔≈│î¶  A╕NDbfAì<ï≈Hì¶⌂H┴Γ♥ §àE☻ Hï↔╞ù☻ Hëâx¶  I;╞u↕  u
DRVSTORE\SamsungRap_91B2624CEE8F4E65E61BE725467EB7EF656E22D2\amd64\SamsungRapidD
iskFltr.sys:Hk╔8Hë∟☺ ☻Hï\$`Hâ─P_├╠╠╠╠╠╠╠╠Hë\WHâ∞`Hï╝$ÿ   Mï╨Hï┘èGI<☺☼ä¢   <♥☼äô
  Hï§*,☻ ïG@ïè╠   ╙α9GD☼ä·   ☼╢OHHïî╩É→   §î╫☺ HïG8HïO0HHë☼╢OHHï♣Ω+☻  îêα§  ☼╢GH
╪+☻ Hïî┴É→   §J╫☺ DïG@ïWDHïO A▒☺Φë  Hâg  E3└AìHï╫Φçè  Θ¿   Hï♣ù+☻ ïê╠   ïG@╙α9GD
☼àì   D;O@uaHï╧Φ╣┤  DïG@ïWDHïO E3╔Φ▓ê  HïW↑Hâg  LìD$0Hï╦Φ▄╢  3╥Hï╧DìBPΦ~▓☺ Hï↔7+
☻ HìW0Hìï╨¶   §▲╫☺ ≡ââα¶  ☺δ&L+W↑Lïä$É   A╙ßI╙ΓHïO ╟D$ ☺   Aï╥Φ♫à  Hï\$pHâ─`_├╠╠
ƒ*☻ Hïë░↕  Φ{Ω■ Hâcx Ç{A t♂Hïâ╕   ÇH♥☺Hï╦Φ¥╟■ <♣u☼Hï♣j*☻ ≡âÇ   3└Hâ─ [├╠╠╠╠╠╠Hë\
    HëH≡HïN@Hï╒HïI►Hï\$0Hïl$8Hït$@Hâ─ _H %"╓☺ ╠╠╠╠╠╠Hë\Hët$►WHâ∞ Hï≥HïÆ╕   Hï∙Hì
;)☻ LcΩïüp¶   ╚à╟t☼ïëp¶  ìA #╟+╚♥∙A╕NDbfï╫3╔Dïτ §┬╓☺ Hï≡Hà└uët$↑WATAUHâ∞ ï∙Hï

This appears to be the possible cuplrit. For now, I have just disabled RAPID mode to see if this eliminates the issue.



FINAL UPDATE: Looks like it's fine now. I looked at this thread here, these fine folks figured out that KB3004394 update (mine updated 12/10/2014, and was labeled RECOMMENDED) may have been the update responsible.

But of course.. that doesn't get rid of the fact that I found actual evidence that there was a leak coming from SamsungRapidD
iskFltr.sys and any kind of leak, regardless of the cause is not good. I certainly don't miss the performance gain for right now, and until this issue is fully resolved I shall keep RAPID mode disabled. (Unless of course I remove KB3004394 as a test and then re-enable RAPID mode, but after days and days of trying to figure out the root cause, I'm just too burnt out to really take it any further.)
 

TRENDING THREADS