HandBrake spontaneously shutting down computer

Support for HandBrake on Linux, Solaris, and other Unix-like platforms
Forum rules
An Activity Log is required for support requests. Please read How-to get an activity log? for details on how and why this should be provided.
fedorahead
Posts: 2
Joined: Wed Apr 08, 2015 2:23 am

HandBrake spontaneously shutting down computer

Post by fedorahead »

Fedora 21 x64 (kernel 3.19.3-200.fc21.x86_64)
AMD FX(tm)-4150 Quad-Core Processor × 4 64-bit
RAM: 7.7 GiB
Graphics: Gallium 0.4 on AMD CYPRESS
Gnome Version: 3.14.2
HandBrake 0.10.1 (x86_64)

Scenario: Used MakeMKV to rip to *.mkv. Using HandBrake to reduce size of original *.mkv. HandBrake spontaneously causing computer to shut down. Activity Log terminates with 'Gtk-Message: GtkDialog mapped without a transient parent. This is discouraged.'

Full log: http://pastebin.com/ahBRUc7r
mduell
Veteran User
Posts: 8187
Joined: Sat Apr 21, 2007 8:54 pm

Re: HandBrake spontaneously shutting down computer

Post by mduell »

Most spontaneous shutdown problems are deficient hardware, usually cooling or power supply.

Code: Select all

Handbrake Version: 0.10.1 (2015030900)
That doesn't look like official HandBrake. Where did you get this build from?
fedorahead
Posts: 2
Joined: Wed Apr 08, 2015 2:23 am

Re: HandBrake spontaneously shutting down computer

Post by fedorahead »

There is no official HandBrake release for Fedora. The HandBrake was installed from http://negativo17.org/repos/fedora-handbrake.repo.

I would normally agree with you that spontaneous shut-downs of computers are commonly caused by one or more hardware issues. However, in this case, the computer only exhibits this shutdown behavior with HandBrake. I have MakeMKV also on the same host and have allowed it to run for hours with no issues. Please understand that I am not saying that this is, specifically, a HandBrake issue. I have a virtually identical computer (only one component, a different brand of SSD OS boot drive, different, and using HandBrake from the same repository, it runs HandBrake just fine. Also, HandBrake ran fine on the host now exhibiting the unexpected behavior until just recently. I apologize for not making my purpose for posting to this forum more clear. I was hoping that someone might have seen this behavior before and might be able to suggest a remedy. Based upon your first response, it would appear the 'GtkDialog mapped without a transient parent' is not a known error of HandBrake. Since I also noticed another anomaly with the host in question, my next step is to reload Fedora 21 and see if the behavior recurs. If you have any other suggestions, I would be happy to receive them, but, if not, please consider this subject closed and thank you for your prompt response.
tlindgren
Bright Spark User
Posts: 260
Joined: Sun May 03, 2009 2:14 pm

Re: HandBrake spontaneously shutting down computer

Post by tlindgren »

fedorahead wrote:I would normally agree with you that spontaneous shut-downs of computers are commonly caused by one or more hardware issues. However, in this case, the computer only exhibits this shutdown behavior with HandBrake. I have MakeMKV also on the same host and have allowed it to run for hours with no issues.
MakeMKV puts very little stress on the machine, while Handbrake stresses it very hard, so this is expected and means absolutely nothing.

If you run it as root and it was an ordered shutdown, it could in theory be the app calling for the shutdown but that doesn't sound like what's happening here.

And if it's NOT, a normal program should not be able to shut down a machine that way which pretty much only leaves hardware or operating system bug/failure!

Pretty much everyone who have these kind of issues and post here doesn't think it's their hardware, but I think so far the record is that pretty much 100% of them has been hardware OR it happens at the end of the encode and they've told Handbrake to shut down at the end (not 100% sure if the Linux GUI offers this, the others do so probably).
mod16
Experienced
Posts: 78
Joined: Sun Jan 12, 2014 11:10 am

Re: HandBrake spontaneously shutting down computer

Post by mod16 »

tlindgren wrote:(not 100% sure if the Linux GUI offers this, the others do so probably).
Yes the Linux GUI does offer a shutdown option. But, of course, only after finishing the queue. According to the log the encoding wasn't finished at all - so I also would guess a hardware/temp/load problem.

To test if this could be the problem I recommend using mprime, which stresses the hardware even more:
http://www.mersenne.org/download/

More info:
https://wiki.archlinux.org/index.php/Stress_Test

Obviously you should monitor the temps before/while running Handbrake or mprime using lm-sensors or something similar.

If the system crashes instantly it's more likely to be a PSU problem, if it takes some time until shutdown, my guess would be a cooling issue.

I guess the Gtk Warning Message is only a consequential error of the actual problem.
trekkie690
Posts: 1
Joined: Mon Apr 20, 2015 11:56 pm

Re: HandBrake spontaneously shutting down computer

Post by trekkie690 »

So, I'm having the same issue with Handbrake-10.1 GUI on Fedora21. I do build from source code (cause couldn't find it per-packaged) so what ever is on the site is what i use. When i do the encoding it only shutdowns when encoding MKV files from makemkv to mp4 (again do this because of size). I have also notice in the journalctl -r that it shows the same information as noted above. As for now im going back to 10.0 until someone can figure out whats causing it, and i can find a .patch file to apply


Just so people know my specs
Radeon HD 7750
AMD FX-8150 8xcore.
16GB-DDR3 ram

P.S. : there was no heat issue with the computer so not HW related
User avatar
s55
HandBrake Team
Posts: 10350
Joined: Sun Dec 24, 2006 1:05 pm

Re: HandBrake spontaneously shutting down computer

Post by s55 »

Heat isn't the only possible hardware issue.

Handbrake can't be causing the issue your seeing, which leaves hardware or something kernel level that's broken on your system.
User avatar
JohnAStebbins
HandBrake Team
Posts: 5712
Joined: Sat Feb 09, 2008 7:21 pm

Re: HandBrake spontaneously shutting down computer

Post by JohnAStebbins »

@tekkle690, I run HandBrake on F21 every day. I'm the linux HandBrake developer and I do all my development on Fedora. I've never seen an improper shutdown while HandBrake is running. So there is something unique to your system that is causing this.
midas
Posts: 11
Joined: Fri Apr 24, 2015 2:45 pm

Re: HandBrake spontaneously shutting down computer

Post by midas »

I too experience spontaneous reboots. It is always while working with H 264 encoding and Blu-Ray MKV files. I have had this for about 6 months and have done some troubleshooting but cannot find any answers.

My hardware is
  • ASRock 980DE3/U3S3 MB
    AMD|8-CORE FX-8350 4.0G 8M R
    2- 8G|KHX KHX1600C10D3B1/8G R DDR3
    XFX Core Edition Radeon R5 230
I'm running Ubuntu 14.04 fully patched. I was running the stable GTK Handbrake and Handbrake CLI but having the same issue with all versions I went to nightly builds. Still no difference. Currently running Stebbins repo at version SVN 7122. I believe it is a derivative of version .10.

When this began, I was running an nVidia GX250 video card. I replaced it thinking it was the culprit as no information is ever in any logs. It is not cleaning shutting down just completely resets within 5% of encoding.

I have run lm-sensors and watched the power and temps. They fluctuate but never hit any critical thresholds, in fact they are quite within reasonable limits. I'll run another and report back on he actuals. I also have run system monitor while watching things progress. Yes the system ramps up nicely :D but then quickly checks out :oops:

Oh, and I replaced the PSU with a Corsair 750 W unit. The case has 3-120mm fans, 4-80mm, and an upgraded CPU thermal cooler (can't remember specifics).

I am willing to do some testing if someone can give me direction. I am really at a loss. I even ran a Windows 7 KVM installation on this server and was able to fully encode MKV files over and over. I realize the CPU architecture is not a complete pass-thru and the video card is not either.

I have found this reported always on AMD CPU's and doing H 264 encoding.
User avatar
JohnAStebbins
HandBrake Team
Posts: 5712
Joined: Sat Feb 09, 2008 7:21 pm

Re: HandBrake spontaneously shutting down computer

Post by JohnAStebbins »

@midas, Try running prime95 and memtest86 to see if either of these turn up any problems. You could have dodgy memory or mobo. Tracking down hardware problems is always a PITA, and can get expensive replacing parts that may or may not be bad :cry:
midas
Posts: 11
Joined: Fri Apr 24, 2015 2:45 pm

Re: HandBrake spontaneously shutting down computer

Post by midas »

Thanks John. It *IS* my CPU :evil: It took a while but the prime95 test crashed it again. Under all other circumstances I do not experience issues, so I guess I have a (mostly) usable data server.

-m
mduell
Veteran User
Posts: 8187
Joined: Sat Apr 21, 2007 8:54 pm

Re: HandBrake spontaneously shutting down computer

Post by mduell »

Try underclocking, IIRC not all the AM3+ mobos can deliver the power the 8350 needs. If underclocking works, a mobo with better power delivery may solve your problem.
midas
Posts: 11
Joined: Fri Apr 24, 2015 2:45 pm

Re: HandBrake spontaneously shutting down computer

Post by midas »

mduell wrote:Try underclocking, IIRC not all the AM3+ mobos can deliver the power the 8350 needs. If underclocking works, a mobo with better power delivery may solve your problem.
Will do. Thanks for the advice.

-m
tlindgren
Bright Spark User
Posts: 260
Joined: Sun May 03, 2009 2:14 pm

Re: HandBrake spontaneously shutting down computer

Post by tlindgren »

It's definitely worth trying underclocking. I see a number of people that have power issues with the FX-8350, several people thinks 6+2 phase power and bulky VRM cooling is absolute minimum for reliable function with FX-8350 and your motherboard looks like 4+1 phase and not much VRM cooling.

Ah, I see that your ASRock 980DE3/U3S3 motherboards CPU support list does include your FX-8350, but has this disclaimer for all the supported FX processors including yours:
* For cooling the CPU and its surrounding components, please install a CPU cooler with a top-down blowing design.

You say you have "an upgraded CPU thermal cooler", often but not always means either Tower type or CLC, both of those violate that requirement which then can lead to overheating of components like the VRM or the chipset that often relies on the "backwash" from the top-down style CPU cooler for sufficient cooling! What are you using as CPU cooler?

My guess all motherboard manufactures had old designs for an earlier AMD cpu and then when the FX series came out they tested it and found that it "works IF certain things is followed". So they just documented the restrictions and then designed a slightly more expensive model with 6+2 or 8+2 phase power and bulkier VRM cooling where special care in setup isn't necessary.

If you have an contactless ("laser") IR thermometer you could take the temp on various component while encoding to confirm that this is the culprit. I assume the VRMs are the row of boxes between CPU socket and the back, the two small heatsinks would also be worth check temperatures on.
midas
Posts: 11
Joined: Fri Apr 24, 2015 2:45 pm

Re: HandBrake spontaneously shutting down computer

Post by midas »

I have tentatively begun the underclocking as a path forward. So far I have dropped from 4.0 GHz to 3.6. I also dropped the CPU voltage slightly too. This failed running Prime95 for about 15 minutes.

Now I'm not experienced with overclocking, etc. So I need a bit more direction. Should I have dropped the CPU voltage? I'm not at my machine right now but there were other chipset voltage changes available too. Any directions are appreciated.

I don't know the name of CPU cooler I am using. I can't find the order on NewEgg or Amazon where I would have purchased it. I believe it is a Cooler Master Hyper 212 Plus.

My approach was to start low and work my way up to get the highest clock speed with the hardware that was stable. Does that sound about right?

TIA,
-m
mduell
Veteran User
Posts: 8187
Joined: Sat Apr 21, 2007 8:54 pm

Re: HandBrake spontaneously shutting down computer

Post by mduell »

I'd just drop the clock at first to find something stable.
tlindgren
Bright Spark User
Posts: 260
Joined: Sun May 03, 2009 2:14 pm

Re: HandBrake spontaneously shutting down computer

Post by tlindgren »

midas wrote:I don't know the name of CPU cooler I am using. I can't find the order on NewEgg or Amazon where I would have purchased it. I believe it is a Cooler Master Hyper 212 Plus.
Like this: http://www.coolermaster.com/cooling/cpu ... -212-plus/

That's a classic tower cooler that provide very good cpu cooling but absolutely no cooling for motherboard components, it's pretty much the exact opposite to what ASRock says is required with a FX-8350 on that motherboard.

I don't think underclocking will be enough to overcome that, at least not unless you go way down in speed and probably voltage too which likely will result in very bad performance.

If it was me I'd probably start with trying to test the theory by removing the side plate and aim a 12" table fan at the motherboard at max speed, if that fixes the crashes you've pretty much confirmed it's one or more of the non-CPU components that is overheating. I'd say 90%+ chance it's the VRMs given what ASRock says.

To fix your issue permanently you need to either replace the cpu, the cooler with a top-down model or replace the motherboard with one that has better auxiliary cooling, usually meaning some kind of "extreme gamer" model.

You could also put in a fan providing airflow over the VRM areas (probably don't need much airflow!), this is the cheapest option. Your chassi might have options for this but probably not and then it requires quite a bit of tinkering to anchor it safely, perhaps you know someone that can help.

Looking at the other options, I took a few example using Newegg for pricing and kept ASRock as the MB provider to limit time spent looking things up.

CPU: You'd need to go all the way back to a Phenom II X6 or lower for it to not have the "needs blow-down cooler". You may be able to find a really cheap second hand 1100T but even that will be a LOT slower than what you have now.

Motherboard replacement: The only ASRock AM3+ motherboard currently is for available that doesn't need a "blow down" cooler for an FX-8350 cpu is the 990FX Extreme9 which cost $170. ASRock is usually among the cheaper motherboard manufacturers so I doubt you'll find something significantly cheaper that can do it either.

CPU Cooler: Blow-down coolers that is rated for 125W or more isn't trivial to find but they do exist, I found three I think really should work well that was also in stock at Newegg (most wasn't rated to handle your 125W cpu)
  • Cooler Master GeminII M4 ($38) - Google lists a number of people are using this with FX-8350's but if you have very tall memory sticks it can have problems, they're all much wider than the cpu and this is the "lowest" of them.
  • Cooler Master GeminII S524 v2 ($44) - Taller which solves the possible memory stick collision but looks like the airflow is fairly directional so it may not airflow where you need it? It probably would work though.
  • Scythe Kabuto II ($55) - Even taller but should provide more even motherboard cooling than the S524 v2 but do make sure it's not too tall for you case, you need at least 1 inch clearance from fan top to case and preferably at least 2 inch.
I'm guessing the GeminII M4 is probably the best choice unless you do have the very tall memory stick issue, CM has detailed schematics that you can check with. If that model doesn't work I'd probably check case height to see if the Scythe works myself. But it's all theoretical on the AMD side so it's just a lot of experience in what should work, if someone has built a AMD FX-8350 machine I'd welcome actual experience!

The $100-$150 CM Hyper 212+ is intended for extreme money-no-object overclocking which explains the price, perhaps you have the far more reasonable Hyper 212 Evo model ($35).
mduell
Veteran User
Posts: 8187
Joined: Sat Apr 21, 2007 8:54 pm

Re: HandBrake spontaneously shutting down computer

Post by mduell »

Even without a blow down cooler, try a fan pointed at the VRM.
tlindgren wrote:The $100-$150 CM Hyper 212+ is intended for extreme money-no-object overclocking which explains the price, perhaps you have the far more reasonable Hyper 212 Evo model ($35).
No, it's a slight variation on the Hyper 212 that came out before the EVO, and it's less than $30. Not at all comparable with high end water or phase-change coolers.
mod16
Experienced
Posts: 78
Joined: Sun Jan 12, 2014 11:10 am

Re: HandBrake spontaneously shutting down computer

Post by mod16 »

I'm also using a Asrock MB (970 Extreme3 R2.0) + 8GB RAM + AMD FX 8350 system mostly for encoding with Handbrake (using Ubuntu/Linux as OS).
Power supply is a Antec "High Current Gamer" with 520W and the CPU-cooler a Thermalright Macho. GPU is a old AMD/Ati HD3850 - but I don't do 3D stuff on this system anyway.

So far everything works fine (no shutdowns/crashes), but at least in the summer months I reached critical CPU temperatures (>60°C), so I upgraded the case cooling a little bit. I placed an additional fan behind the MB/CPU-socket and at the front "intake".

Also I undervolted the CPU (without reducing the clock - still running at 4 GHz) and tweaked the BIOS/EFI setup a little bit. Here is what I wrote for myself down after doing some hours of research on how to optimize the setup:

Spread Spectrum --> disabled (not really important)
AMD Turbo Core --> disabled
Load Line Calibration (LLC) --> disabled (reduces load on the poor voltage regulators)
AMD Application Power Profile --> disabled
Thermal Throttling --> disabled (be careful with that one!)
VCore --> 1.265V (1.250V was also stable, but I decided not to overdo things; below that mprime produced errors)

This CPU is some kind of a wild beast you have to tame first. ;-) But after that it's quite good, especially for video encoding.
midas
Posts: 11
Joined: Fri Apr 24, 2015 2:45 pm

Re: HandBrake spontaneously shutting down computer

Post by midas »

So, I first went and bought an 8" fan. I have the case open and blowing highs peed directly down on CPU and chips. It made no difference. Crash. I then took the clock speed down to 3.0GHz. Runs fine. lm-sensors reports the k10 CPU temp below the default high of 158f. Of course that is way below where I would like to run. I bumped the clock to 3.6GHz and saw the temp spike quickly over 178+ so stopped the Prime95 tests. I then dropped it to 3.4GHz and saw the temp stabilize at around 178f.

Two things:
1) from what I have read, lm-sensors doesn't really report the actual temp of AMD CPU's. I don't know where it got a default number of 158f being okay but I'm 20 degrees over that but running now for well over 2 hours.

2) The temperature reported moves to quickly to be accurate. When ramping up or stopping Prime95 the reported temp changes 20+ degrees within just a few seconds. Ramping up I could see this but heat dissipation doesn't look right.

mod16- Thanks for the info. I'll try it out. How different is your MB from mine? What voltage did you drop to? [:Sigh:]. Looks like some more research...

-m
midas
Posts: 11
Joined: Fri Apr 24, 2015 2:45 pm

Re: HandBrake spontaneously shutting down computer

Post by midas »

***UPDATE****
Crashed just under 4 hours... :?
****************

Just some more info. I'm at over 3 1/2 hours with the clock at 3.4GHz (too slow :x ). Here is what my sensors -f shows:

Code: Select all

radeon-pci-0100
Adapter: PCI adapter
temp1:        +99.5°F  (crit = +248.0°F, hyst = +194.0°F)

k10temp-pci-00c3
Adapter: PCI adapter
temp1:       +172.6°F  (high = +158.0°F)

fam15h_power-pci-00c4
Adapter: PCI adapter
power1:        6.75 W  (crit = 125.19 W)

nct6776-isa-0290
Adapter: ISA adapter
Vcore:          +1.38 V  (min =  +0.00 V, max =  +1.74 V)
in1:            +1.86 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
AVCC:           +3.31 V  (min =  +2.98 V, max =  +3.63 V)
+3.3V:          +3.31 V  (min =  +2.98 V, max =  +3.63 V)
in4:            +2.04 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in5:            +1.70 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in6:            +2.04 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
3VSB:           +3.46 V  (min =  +2.98 V, max =  +3.63 V)
Vbat:           +3.36 V  (min =  +2.70 V, max =  +3.63 V)
fan1:             0 RPM  (min =    0 RPM)
fan2:          2198 RPM  (min =    0 RPM)
fan3:             0 RPM  (min =    0 RPM)
SYSTIN:         +93.2°F  (high = +32.0°F, hyst = +32.0°F)  ALARM  sensor = thermistor
CPUTIN:        +200.3°F  (high = +176.0°F, hyst = +167.0°F)  ALARM  sensor = thermistor
AUXTIN:         -80.5°F  (high = +213.8°F, hyst = +213.8°F)  sensor = thermistor
PCH_CHIP_TEMP:  +32.0°F  
PCH_CPU_TEMP:   +32.0°F  
PCH_MCH_TEMP:   +32.0°F  
intrusion0:    ALARM
intrusion1:    ALARM
beep_enable:   disabled
Last edited by midas on Wed Apr 29, 2015 6:08 pm, edited 1 time in total.
User avatar
s55
HandBrake Team
Posts: 10350
Joined: Sun Dec 24, 2006 1:05 pm

Re: HandBrake spontaneously shutting down computer

Post by s55 »

CPUTIN: +200.3°F (high = +176.0°F, hyst = +167.0°F) ALARM sensor = thermistor
That's right on the limits. I'd never want to see a machine running that hot personally.
Woodstock
Veteran User
Posts: 4614
Joined: Tue Aug 27, 2013 6:39 am

Re: HandBrake spontaneously shutting down computer

Post by Woodstock »

Wow, I get worried when I see 150F on mine, and turn a second fan on...
midas
Posts: 11
Joined: Fri Apr 24, 2015 2:45 pm

Re: HandBrake spontaneously shutting down computer

Post by midas »

As I've said, what I can find says the numbers reported are suspect for AMD CPU's. And I have a jet turbine worth of air flowing through the box. In further researching, the numbers reported can be very, well, wrong.

FYI:
CPUTIN = CPU Temperature Index from the motherboard, not the actual CPU
AUXTIN = Auxiliary Temperature Index
SYSTIN = System Temperature Index


I'm really at a loss but will try mod16's suggestions in BIOS. Aside from something flakey in the hardware, I can't see why I am seeing these numbers (assuming they are correct).

-m
midas
Posts: 11
Joined: Fri Apr 24, 2015 2:45 pm

Re: HandBrake spontaneously shutting down computer

Post by midas »

Google is your friend. From reddit:
Hi, topic basically says it all, CPU overheats on load, shutting down the PC or thermal trottling when socket temp goes over 65°C.

Googling taught me that the problem is due to bad power design of the mainboard. The Asrock CPU support list tells me to install top down cooler (got a tower one currently, Skythe Katana 4, also tried an AC Freezer 13): http://www.asrock.com/mb/AMD/980DE3U3S3/?cat=CPU

Things I also tried: New BIOS, Additional casefans, Stock cooler (top down), Lower Voltage,

New board or new cooler? I think a new board, can someone recommend me a board that guaranteed has no such trouble which is not too expensive aswell?
Sooooo, I may be in the market for a new MB. I'll try mod16's suggestions first. Well, when I get home and shutdown the crashed box.

Just to hedge my bets, any MB suggestions?

-m
Post Reply