Computer locks up linux, after 10 minutes, hour, 2 days. 2 and half max!
 

My computer is pretty new-- AMD Athlon XP 2500 / 512megs of 3200 ram / 2 Xbox Hard drives 10gig, and 7.4 gig / ASUS A7V600-X series MoBo. / eVGA GeForce FX 5500 256 / Antec AiroFlow CPU Fan / a decent case. / and Slackware 10 2.6.7 kernel/

Now on to the problem. My computer just.. LOCKS UP on me randomly, at random times. Usualy, and more oftenly when im playing a game. Like DoD, Hostil Intent, Doom3 (demo), Battlefield 1942, Desert Combat << o god! This is the worst!, The SIMS! << nvmd about DC, this is the worst. And if im not playing a game, well then, the next day, if I leave it on running all night trying to download something, I wonder about on the desktop for a little bit (tiny tiny bit) and it gives a hard ass lock up. Ill be luky if it makes it to two days. Thats if I do absolutely nothing, or it just feels like giving me an extra day.

Then I restart and all is good again. Till it strikes again ^^^^

I've been lately trying to find this damn problem. This is what information I've been getting from my tests... (plz plz read from this point forward, if u havent been)

1. Check if ram was bad. Looked at it. Nothing unusal, nothing somewhere where it shouldn't be. I did memtest, for 40 f'n hours, and I finaly got the go ahead from someone to stop, since it didn't have any errors.

2. Mobo, I havent check, I dont know what to check, even if ther is something to check, I wouldn't know where to begin. I highly doubt this is the problem or my computer wouldn't turn on period. And I would've punched it. (i done it before, ill do it again)

3. Processor seemed like it was overheating, but from an overwhelming amount of ppl, they say 140 degrees is average for an AMD. So I agreed. And I noticed my friends brother's AMD is the same, so I moved forward. Also I got a new fan, and I put ENOUGH thermal greese on. (not overwhelming, but enough to know I put enough)

4. Power supply is brand new, 350 watt, and it works, mebe. I've got it from a place I've been going to for a while, and they havent givin me any problems, so I know this isn't the case. Or like before, the shit wouldn't turn on. So we move forward.

5. I checked how much ram my Slack box ware is using. Its a decent amount. I use fluxbox now, and it still stays around the same amount of ram-- 75-350 megs, sometimes the whole damn stick, but my sources (some of you guys) say that taking up the whole thing is the way Linux works. I'll have to go with that for now.

6. My graphics card, and case -- I sumed these 2 up because without a reasonable doubt, I assumed that these little guy wouldn't be the problem. Mebe. Actualy if you have a DaMn good reason why it could be one of these two, I'd like to hear it. I'd realy realy like to hear, so plz dont be the fat kid in the class and be afraid to speak up.

7. My hard drives.. ok ok... I know this shit is been used and abused, and if ther was a law against abuse on hard drives, I'd get the death penalty. But these seemed to work JUST FINE before, so ... I dont know. But I talked to this tech guys at the store I go to, and they said my shit writes all the time, even when the light doesn't blink. It writes, like my personal slave. (and master :confused: ) so I believe these little guys aren't doing what ther supposed to, and they stop writing during the time im on the computer. So, could someone back this up for me. I believe this is the source of the problem, I can't prove it, so I failed my case. Plz, again, speak up. I need to hear opinions! I'd rather get this box fixed, rather then seeing my fists marks on it again.

8. Linux itself.. YA RIGHT! Na jk. Maybe, I dont know, i've gotten soo despret that I've turned against my buddy. Plz let this be not the problem.

9. I fill this in as I go. The above is pretty much what I think could be the problem. So plz, give me any opinion, no matter how far fetched it sounds. (dont take that seriously)

Form what you describe, the psu is the most likely suspect IMHO. If it's a good brand(antec, thermaltake, etc) 350W should be OK. If it's some no name piece of crap, I'd be very suspicious. Your wrong thinking psus are digital, i.e. they either work or they don't. It's just not the case. Poor quality psus can give inconsistent power; when the voltages go too far out of spec your computer can shut down or cause all kinds of weird problems. I learned the hard way to never ever try to save a few bucks with a cheap psu; it's just not worth it. Here's a link rating some brand name psus:

http://forum.pcmech.com/showthread.php?t=98650

Well, I did a test on how much wattage I need. It said I needed a 283 watt PSU. I have a 350. But its listed under bad, its an L & C. But I don't want to go out, and buy Power Supply and realize that thats not the problem. How would I go about to figure this pickle out? I got an idea, I think Ill try it, but its a shoot and miss. And my target is far away.

Punching the motherboard does not do anything. It creates more problems then fixes. Some traces are very, very narrow. Heat and bending can ruin a motherboard. Transporting a computer makes the matters worst because of ion build up and bending.

I do not care what anybody says. 140 degrees F is way to hot for any semiconductor. Do everything you can to reduce it to something like 120 or less. Never assume temperature ratings are true. You can try to increase the voltage by 0.1 volts for the processor.

I recommend power supplies from Enermax, Zalman, Power & Cooling, and Antec. Antec is the lowest I would go for power supplies. Power & Cooling has the best power supplies to date. Power & Cooling under spec their power supplies, so this means they are heavy and they can output some serious power. Enermax and Zalman are great for moders and quiet areas.

An In-line UPS and a surge protector should be used. Again make sure the UPS is an in-line type not stand-by type.

If you can set AGP to 2X or even better 1X, you will get better stablity.

Could be cron and sylog running. You can stop them to see if it the lights stops blinking.

It could be Windows. Those games that you mention are written for Windows. If you are using WINE, it could be WINE just freezing X Window Server. SSH into your computer to terminate X Window Server and bring it back up. If you see your keyboard blinking then something software related went wrong.

Computer power supplies are digital. The type of power supply that computers use is switch-mode power supply. It uses a combination of frequency generator, comparater, a transistor or MOSFET, diode, inductor, and a capacitor to regulate the DC power supply. Switch-mode power supplies are very, very noisy than linear power supplies on terms of ripple voltage. Switch-mode power supplies are able to step-up, step-down, and invert the voltage. Also they are very, very efficient than linear power supplies. Linear power supplies can only be used to step-down the voltage and they are very poor using power. Most of the power in linear power supplies is given up as heat.

The lower ripple voltage the better the power supply. Usually ripple voltage is stated in millivolts (mv). A millivolt is 1/1000 of a volt.

I'm still only moderately experienced, but I can add to all of the above:

(1) I would suspect that hot processor, too. I had a PC behave in _exactly_ the same way, and it turned out that the processor's fan was slowly going out.

(2) Lacking that, don't be afraid to get a new power box. Just save the reciept, and if it doesn't turn out to be the solution, just take it back. I do that with CompUSA all the time.

(3) Always easier to hook a suspected powerbox into a similar computer, to test it. If you have another system, that is!

(4) Ignoring hardware for a minute, and assuming the processor is fine, I would then guess (dimly) at a memory leak from some program or another. Have you checked into softyware fixes on some background process or daemon?

humbly offering what little I've got...
Hosiah

what would that do? wouldn't that increase heat? Sry for asking many questions, im learning at the same time as I am fixing this problem.

Forgot about that. When I gets some cash, i'll have to go do that.

So far thanx for the response! They seem promising!

I agree with Electro - that temp is way too hot for a CPU. I've got the exact same chip and mine ususally runs at about 115 F. Are your fans running? Are the installed correctly (so that the airflow is all in one direction)? To illustrate, you would not want the front fan and the back fan to both be blowing in. Instead, you should have the front fan blow in and the back fan blow out. Refer to your cabinet's user manual for info on fan placement.

Also, are you overclocking? If so, that's a prime source of instability. Back it back down to a normal speed. Good luck with it -- J.W.

No, I tried overclocking once but it didn't work. So I changed it back. I dont have any fans inside anymore, I took off the front panel, and I keep a fan close by, to keep air flowing, but I dont have it directly hitting the case, my friend told me it messes up the hard drives.

Your friend is incorrect. Fans do not mess up hard drives. Fans cool off the hard drives. I have two fans in the front as intake that are in front of my hard drive cage. The fans cools off my hard drives when they are writting or reading for several hours a day (video capturing). Very hot hard drives do not last long as cool hard drives. You should have atleast a fan in the back to take air out of the case. This does not include the power supply fan. AMD and Intel recommends you do this for sufficient air flow. A fan on top of the case (optional) helps get rid of heat from CD-ROM, DVD-ROM, or any 5.25/3.5 inch device.

That I doubt. The temp could be it, I have an Athlon with the Palomino core (the one that gets really hot really fast), and right now it's at 118F, gets to around 130F when it gets to CPU-intensive tasks (compiling, DooM3, etc). I have one fan on front blowing things in and the back blowing out with the CPU fan running fine, and I haven't had any HDD problems, so put those fans in and see what happens.

PS: I don't know about 140F being a bad point for AMDs. When I had overheating problems, the thing was at 160F.

At the risk of repeating what's already been said, your friend is misinformed. Keeping your machine running at a cool temperature is essential, and I'm not exactly sure how you could possibly "mess up" a hard drive just because air was flowing past it.

Indeed, forcing your drives to operate at higher than normal temperatures will likely shorten their useful service life, and could endanger your data. In short, buy a couple of fans for your case. That should cause the observed temp to go down, and likewise the heat-related instability issue should disappear as well. -- J.W.

So far changing my AGP speed to 4x has made my system a hell of a lot more stable. I've been letting my computer run for about a day and a half without any problems so far. Thanx you guys! :D If ther seems to be any more problems, I post back.

Changing the AGP rate did nothing. It was just a matter of time before it went belly up again. What steps could I take to cool down my damn processor. Its ALWAYS at 140 degrees or more. Im starting to dought ppl with 140 degree processor that has a stable computer. Plz reply.

Look up water cooling and vapor chill.

I completely agree with all of you, but i think you're focusing too much on hardware issue (that could be perfectly), but i could be as well Slackware (module conflicts??, bad kernel config??). Does it hangs as well in Windows?? If so, it's clear that hadrware related problem, but if not, i would point to change distro (fc2 for example).

Defeitely 140F degrees is hell itself for a cpu, and could be the problem. Cool that cpu ASAP (thermaltake has several solutions to that: www.thermaltake.com).

Hope you're gonna make it ;)

In my experience 140F (60C) is not hot enough to cause failure. AMD processors run hot and that's about what you can expect from the stock fan. Crashing while running 3D apps incdicates that the video card doesn't have sufficient cooling. So you may have a flacky VGA card or a bad fan cooliing it. Also power supplies can cause your computer to crash or create weird problems even though your computer boots up and seems to run fine. It wouldn't hurt to go buy a $20 350W supply to try and just keep on hand in case of a failure in the future.

I have a 350 watt power supply, but like that one person posted, it was a bad brand. Ima look into buying another. The processor runs hot, even tho I bought another CPU Fan. (areo flow) I took off the side case, so Ill have complete airflow other then having it on. For that guy that said to try FC2, I might, because in the year I have used linux. Only the Red Hat products never gave me a problem. Red Hat 9, Fedora Core 1, Fedora Core 2. The only reason I switched to slackware or others, is either 1. to try different flavors 2. to go with a harder distrobution. I don't have ANY money right now, so I can't right away go and buy a new power supply. I want to know of a free way to cool down my computer, EVEN by 10 degrees. 130 is better then 140 degrees.

:lol:

;)

That's funny.

Anyway, yes CPU temp can be related to your problems...

Now, I will say that 140F is fine and not to worry. However, like most motherboards, the temp diode may be off by a few points.. (sometimes up to 15C)... so your temp might even be higher then 140F in which case I would look into better cooling.

140F is around 60C (I think..) and that's just about where I draw the line. My home computer's CPU (Athlon XP 1600+) runs anywhere between 45-50 idle, and 55-58 load. Some might find this to high.. but until my computer locks up, which it has yet to do, it works just fine for me (I prefer not hearing my CPU fan, it's silent at 1300rpm).

Face it. Cooling your computer is important... but there is no need for extreme cooling. An Athlon chip will run just fine even at 65C (or 70C for that matter). It's within the chips thermal specs. The problem is with the motherboard... when a chip runs at 70C some of the heat travals through the PCB (which is a method of cooling some chips and parts mind you) and this heat will cause instability within the chipset or supporting chips.

The bottom line. If you think that heat is the problem... invest in better airflow within your case, not just over your CPU.

That aside...

I think it might be a buggy driver.. you never know.
I would start with a bare kernel with no support for anything... then modprobe device by device putting it through it's paces on each addition and seeing if it locks up. If it does, play around with the drivers that you have loaded until you root out the buggy one... then look for a replacement or another way to use the hardware. Oh, and do this while not in X. . . and load X if the computer runs fine without it... and note if it locks up.

Video drivers are a bitch... I'll tell you that much.

Ill first try and get the thing cooled down. Does anyone know of a command I could use to see the temperature of my comp. Other then restarting it to get into bios?

cat /proc/acpi/thermal_zone/THRM/temperature

I dont have thermal_zone, do u know of a program?

Install and setup lm_sensors. lm_sensors are off too so do not depend on it too much.

slackMeUp, I was not joking. I was serious. 140 degrees F (60 degrees C) takes a lot of work or power to take the heat away. Like I said earlier. Any semiconductor can not handle heat very well.

K I think I install them. The readme says only to do "make user" and "make user_install" for my 2.6 kernel. Then I can't find the program.

--EDIT--
nvmd. I did a modprobe i2c-proc, and then I did "sensors" but it said "no sensors found"

How about those little applets that you see on fluxbox and stuff, that tell you your hard drive activity how much ram u used, and the heat of your processor. Sry I dont know what ther called.

So far thanx for the help.

I put a fan near my computer, and im cold, but when I play my games, every so often it starts to have a lot of harddrive activity. But it doesn't lock up, this is where I think the computer would overheat because of all the hard drive activity, but now since its very very cold, (im not exactly sure, thats why I need a temp gauge) it doesn't seem to lock up on these hard hit hard drive activities. I havent given a full on hard course test yet, I'll reply back when I do so.

BUMP!

My computer is running solid 112 degrees, but I cant play any game without a lock up. (native or cedega) I can't check to see if my modules are the problem since 1. I dont know where to begin. 2. I have never compiled a kernel successfully, I think its just stupid. I have been thinking it Slackware now. Would Fedora Core 3 or other distroubtion solve this problem?

Well, it's not your processor. That leaves either faulty motherboard, faulty video card/settings, or something in Linux that must be fixed. You sure the video card is alright? Is Fast Writes or SBA on? What's the AGP Rate?

AGP rate is 4x, SBA Fast Write I dont know. I forgot the command, but ill try and google it later on today.