Random rebooting

[BOXXlabs]

"...What could be reason for such behaviour?"



"dirty" power (brown-outs or power fluctuations)
a failing power supply
a failing component
a failing motherboard

It's more than likely one of the first two. It's not likely to be the second two, but If you replace the power supply and put your system on a UPS then it's gotta be one of the cards or other system components.

Good Luck.

[Cartesius]

Had the same problem not long ago. My computer rebooted sometime during the night and I couldn't figure out why until one day it hit me: it was overheating. I had it placed on top a small desk and during the mornings the sun shone right on it and just fried it. As soon as I moved the computer away from the direct sunlight the rebooting stopped. Just to be sure I now run it with the cover removed and an ordinary desktop fan cooling it.

/Anders

[biliousfrog]

Make sure that the system isn't set to reboot automatically on system failure, otherwise you'll not get a BSOD which could tell you a lot about any simple issues such as problem drivers. You'll kick yourself if you end up replacing a bunch of components and the problem was due to a driver or software.

System - Advanced - Startup and Recovery - Settings - System Failure - make sure that "Automatically Restart" is unchecked.

[ienrdna]

Thank you all for responses

biliousfrog (http://forums.cgsociety.org/member.php?u=151597) I did, lets see what happens. I do have failing gfx card, sometimes images turn black or white in browser, videos tend to lock up whole system for some time, and pretty much all apps that depends on directX crash. Didin't play eny games for year or so.:cry: Guess it time for new gfx card.

ps. Eny suggestions where I could find quick replacement for geforce 7900 gtx?

[Buexe]

maybe it is a heating problem? is there enough air flow going through the system? I had a computer where the whole rear was blocked with dust, only realized it when the computer started to act strange.

[BOXXlabs]

yep - could be heat...

Make sure all your fans are spinning and the air vents are clear...

[imashination]

With all those symptoms id bet money on it being heat. Remove all the dust you can from the metal heatsinks and check all the fans are spinning.

[ienrdna]

Hmm, it could be. But when something is rendering, fans start screaming, but rebooting never happens during rendering. Also usually it happens during night and and sun is shining on to compurer literally all day long.

[biliousfrog]

Have you installed any new software or hardware recently? Is there something running in the background when it reboots, perhaps your antivirus kicks in at night or something else that's causing it to crash?

I'm not dismissing a hardware problem but if it always occurs at the same time perhaps there's a process starting and causing a stop error...which will result in a BSOD now that auto-reboot is switched off. I can't run Media player because my audio driver conflicts and causes a stop error for some reason...the only solution is to not use media player!

Something else that I've thought of....Windows updates! My PC rebooted mid-render yesterday because Windows updated itself and I wasn't there to cancel the restart. Be sure to turn anything like that off if you're leaving it unattended.

[DigitalBlaspheme]

Well, recently my computer started robooting randomly. I lost couple of renders this way, and I am getting pissed of.
What could be reason for such behaviour?Always keep in mind that there are, at the very least, 2 things a computer will NEVER do:


1. They never make mistakes - the person that coded the program/put together the pc made the mistake. It is theoretically impossible for a computer that is programmed to follow 'orders' to make a mistake with the given orders, a computer isn't amelia bedilia...but it very could be as it acts just like this literal maid. They do EXACTLY what they're told to do, they can't say oh well that wont work lets do it this way, unless its in the code for it to do so.


2. They never do things 'randomly' - it's just not in the code, even if it IS in the code. Its another one of those theoretical impossibilities. Understand that? No....don't worry you're not the only one. Let me explain - There is always a reason for the madness....it's just a matter of finding out what is causing these crashes i.e. is it because of faulty power, an overheating cpu/memory, or is it a program issue. There are other issues that can cause this but these three are the.......uh.....most prominent. Just remember when it comes to computers even a program to generate random numbers is NEVER random. Randomness can be faked though, like when some companies generate thier keys for thier programs....however it's ony faked...there is ALWAYS a method to the madness. But that's not to say that the action wouldn't 'appear' to be random.....it's to say that the action ISN'T random...no matter how it appears.


Lets see now, I've been working on computers for about 15 years now and I've yet to lay hands on a computer that succeeded at either of these 2 impossibilites. If you have one that can....send it over to me and we'll both make a pretty penny if I can prove it.




I know this post wasn't very informative as to what could be the issue, but I've seen enough answers that I agree with already posted so I figured I'd post this little rant.

[ienrdna]

DigitalBlaspheme (http://forums.cgsociety.org/member.php?u=314307) Thanks, I understand all of that.

I need to specify something. Rebooting never happens all eny specific time, just mostly at night and never during rendering. Only new hardware I istalled was wacom tablet( could that mean thers less energy for main system?).

[BOXXlabs]

ok - so:

1 - it appears to happen "randomly" (and symptoms DO manifest "randomly" when components fail) and can't be reliably repeated by doing anything in software - so it's probably not a software issue.

2 - and it doesn't happen during cpu-intensive tasks like rendering - so it's probably not heat-related

my best-guess is that it is either a VRM (voltage regulation module) issue on the motherboard - or a failing power supply.

My advice is to go buy a new power supply for $50 - it's the cheapest and quickest (and most likely) thing to check first.

[ThE_JacO]

Lets see now, I've been working on computers for about 15 years now and I've yet to lay hands on a computer that succeeded at either of these 2 impossibilites. If you have one that can....send it over to me and we'll both make a pretty penny if I can prove it.

You clearly never had a pentium then:
http://en.wikipedia.org/wiki/Pentium_FDIV_bug

:)
It's actually perfectly possible, in practical terms, although very unlikely, for a computer to fail for reasons other than programming or manufacturing issues.

Cosmic radiations altering a bit in transit (99% addressed by ECC ram), high levels of static charge in the air, channel desync etc.

If we want to go into philosophical realms, which your post roams freely, you must have missed out quite a lot in those 15 years, and plenty engineers designing chips day in and day out would agree that you did.
The ghost in the machine is more than a myth ;)

As for the actual post: I'm placing bets on the PSU first, and on a videocard 2nd (those are usually left alone while rendering but do suffer from overheating if an OGL screensaver kicks in)

[DigitalBlaspheme]

If we want to go into philosophical realms, which your post roams freely, you must have missed out quite a lot in those 15 years, and plenty engineers designing chips day in and day out would agree that you did. philosophical? hardly.


The ghost in the machine is more than a myth ;) Yeah, and so are the gremlins that steal my socks out of the dryer, who's being philosophical now?


You provided a link to a bug as proof that a computer can be incorrect. A bug is what? Say it with me.....human error. But since you like wiki's so much HERE (http://en.wikipedia.org/wiki/Software_bug) is a link you'd probably listen to before me. You'll probably say, but what its dealing with software bugs. There is no difference between the two. One is with written instruction, and the other is with hardware....both types of bugs produce the same effect. Undesirable results. You can't possibly think that if I build a computer that works for....well lets sday 5 months with no issues, and i take if offline effectively making it unupdatable, and turn it off for 5 more months, that it will pop up with a software bug on the first bootup? A hardware malfunction, sure maybe....but a bug no. And a hardware flaw (dont confuse this with malfunction) has the same chances of 'popping' up as the software bug. Both types of bugs have a cause. The trick is learning, not only how to identify them, but how to repair them. which is why there are certain credentials that are needed for these some of these positions, the rest teach you on the job. Noone wakes up and says, "wow I learned ohms law in my sleep, or wow now I understand signal flow, or wow I can go work for NASA as a rocket scientist, I shoulda slept longer when I was a kid".


nothing a computer does is random, it is all pre programmed. If you could create a computer that could think on its own then maybe i'd get behind you and say a computer could do random things. but they don't. Even the BEST AI out today follows preset paths. Here's a test that nothing is random......and yes ill get philosophical for you. Clear you mind and then try to do something random. Anything. Anything at all. What'd you do? You jumped around like a retard didn't you? That's my guess, but it probably wouldn't be right. However, your random act.....was thought of by your subconsious mind before you did it....hell you probably thought of 10 random acts, and here you thought the one you did was random. You'd be surprised to find out how much faster you thought of doing something before you did it, random or otherwise. The only time you would do something UNPREDICTABLE (big differnece between that and random) is if you tripped over something you didn't see. You'd operate on reaction alone.....or so you'd think. You thought about doing that reaction and more long before you started falling forward.



NOTHING at all that has a CAUSE is random. So far you've all given him probable causes in the form of your ideas on what it is. Some have even mentioned specific components. How you could have done that if this issue was truelly random? HERE'S (http://en.wikipedia.org/wiki/Random) a link to the wiki definition of random seeing as you like wiki articles.


Sure you may have a piece of equipment that seems to die on you for no reason, nor you or the best buy geek can find a cause. There is one. And if its an issue with the hardware, and RMA will get you a new one. However, I'd have to guess that at least 7 out of 10 (note i haven't researched this is just an educated guess) of all RMA'd hardware is refurbashed and resold by the manufacturer. I did it for 3 years with transformers and I doubt a transfomer is the only piece of hardware that can be fixed. And waht would they need to do in order to refurbash your hardware. Find the cause of malfunction. What did the wiki say? Anything with a cause, among other things, CANNOT be random.


Nothing man does is random, nothing man is capable of creating, prior to this writing, is random. While it may appear that there is no difference between random and unpredictable, the two only share one commonality, to someone that can't make the distinction between the two, they both have the same results.

If we want to go into philosophical realms, which your post roams freely, you must have missed out quite a lot in those 15 years, and plenty engineers designing chips day in and day out would agree that you did. The difference between me, with my 15 years experience, the engineers that design these chips and you is that, with the right tools, we'd diagnose and repair the issue, where you would probably have to go to someone like me or get the hardware RMA'd. But that's just a difference in education and knowledge of the hardware. In the future, prior to critizing someone's expertise, do some research on them or the facts you're argueing about. linking a bug to show that computers ARE fallable lol, nope sorry the processor was handling the calculations the way it was made to.....to bad the way it was made to handle the calculations didn't produce the correct calculations. Computers don't just say, "I know 2+2=4, but **** this assholes been running me ragged for the past 8 months, I'm going to tell him its 5." But, if its in the code that the math rule to follow was SADMEP instead of the correct PEMDAS, then sure....you'd think the computer was fallable, and you'd link the same link you did in order to prove it. Computers are infallible, the code used to get them to do what they were made to do isn't....its as fallible as you, me, Tom, Dick, and Harry.

[BOXXlabs]

It's an academic argument - completely philosophical.

As an electrical device (let's say a transistor) fails, the circuits (let's say 'junctions" or "gates") carrying electrons become unstable.

Perhaps at the atomic scale, if we were able to observe and analyze the failure mode, we might be able to quantify the extremely minute fluctuations to the immediate environment that could cause electrons to flow or not flow from one point in time to the next. I suppose that one could then explain exactly what was happening and the exact nature of the conditions that cause the intermittent, problematic behavior.

But from our macroscopic POV, the behavior of a failing electrical transistor or device can and will manifest itself as RANDOM.

done. over. fin.

coffee anyone?

[RPG2006]

What's the exact spec of your machine? very much a guessing game without knowing that. Is it overclocked or running stock?

If the PSU isn't man enough for the system, then it could have possibly done some damage. I know my son's graphics card died thanks to a wimpy PSU.

If it happens at night, could it possibly be the fact it's been on all day, beforehand. If you're not rendering at night, dare I ask what are you doing? Just looking for a pattern. i.e. watching videos or playing games.

As someelse suggested would be good to see what the bluescreen message is.

It's very hard to give an exact diagnosis. It helps to have the facility of having spare hardware at hand, but I think the point someone else made about not just rushing out and buying hardware is a good one.

Maybe an idea to run some stability tests on your kit.

Download http://www.alcpu.com/CoreTemp/

Download Prime95 (A torture test application) http://www.majorgeeks.com/Prime95_d4363.html

Download cpu-z

Run Coretemp and Cpu-z. You can monitor your cpu temps and your cpu's voltage.

Then keeping them both open run Prime. Run the blend test to test your memory. You want to leave this atleast for an hour or so. Run the small FTT test, to test your cpu, again for a few hours. Monitor the temps and voltages while you do this. If you see crazy temperatures and it bluescreens that would certainly highlight one possiblity. If it bluescreens in Blend pretty quickly, but not with small FTT, then that would point to memory or the motherboard being a problem

If these pass with flying colours however, then you can eliminate them (to a certain degree) from the cause of your problems.

Further tests you can try are OCCT ( A very intensive torture test). You run this for an hour. Another one Memtest, which you run from DOS.

Certainly an idea to get everything backed up I'd say. In addition from a graphics card point of view maybe worth downloading 3dmark06 and giving that a few loops. What temps are you getting on your graphics card and what fan speed is it at? Download rivatuner from www.guru3d.com if you want to monitor this.

Another thing you have to bare in mind is if your machine is unstable and has bluescreening, then there's a good chance your OS is now corrupted as well.

RPG

[LRomero]

Well, recently my computer started robooting randomly. I lost couple of renders this way, and I am getting pissed of.
What could be reason for such behaviour?

I suggest doing a memtest to make sure your memory isn't failing. Start making full backups of your drives as it may or may not be failing. Random restarts can be a cause of many different things so just have to diagnose it right.

[Cartesius]

Well, it looks like my theory on overheating was incorrect regarding my own random reboots:

http://forums.cgsociety.org/showthread.php?f=23&t=658179

/Anders

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.