Message boards :
Number crunching :
Computer spontaneously restarts when running BOINC?
Message board moderation
Author | Message |
---|---|
Scooley01 Send message Joined: 25 Apr 03 Posts: 45 Credit: 259,298 RAC: 0 |
I'm not sure if anyone else has had this problem, but lately (only the past couple weeks) my computer has been randomly restarting. After playing around with which programs were running, I discovered it only happens when SETI is running. So far, I haven't seen any damage from the restarts, but it is a sort of inconvenience that I can't leave my computer unattended without losing anything that was open. Has anyone else experienced this problem? I upgraded BOINC, and to no avail; the problem remains. --- Computers Crunching: Laptop (1.4GHz, 512mb RAM) Main Laptop (Dual Core 2.16GHz, 2gb RAM) |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
> Has anyone else experienced this problem? I upgraded BOINC, and to no avail; > the problem remains. > I've seen my laptop(S) do this. The Seti App runs the CPU full time. This creates an increase in heat. This OVER heating causes my system to hard reboot. The problem wasn't seti/boinc's issue. Have you cleaned out the fuzz from your heatsink? Do you have Temp sensors you can monitor? hope this helps tony |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 |
SETI is certainly going to run your system hard. Spontaneous reboots like you report are genereally hardware failures, or duff device drivers. Unless you've added any new hardware recently or upgraded drivers, the hardware is the most likely cause. The components most stressed will be your CPU and Memory. Possible CPUs problems are overheating or poor power supply. Memory problems would be poor power, faulty memory or just slightly 'out-of-spec' memory. You dont give any hardware/software spec for your system so I'll guess at Intel or AMD and some flavor of Windows.... I'd grab a motherboard monitor (google for 'MBM5') and monitor the CPU Temp and voltages if possible. Also grab a copy of MEMTEST86+ (yes... google for 'MEMTEST86+') Between them you should get a good idea if there are any hardware issues you need look at. |
Chilean Send message Joined: 6 Apr 03 Posts: 498 Credit: 3,200,504 RAC: 0 |
|
Scooley01 Send message Joined: 25 Apr 03 Posts: 45 Credit: 259,298 RAC: 0 |
I am almost 100% sure it isn't hardware, as the hardware on this has been installed and configured the same for 3 months. I'm also fairly certain it isn't overheating, because I have an awesome heatsink with two fans on it, and 2 more case fans. Possible, but not likely, in my opinion, however, just to be sure, I am installing MBM5. --- Computers Crunching: Laptop (1.4GHz, 512mb RAM) Main Laptop (Dual Core 2.16GHz, 2gb RAM) |
Scooley01 Send message Joined: 25 Apr 03 Posts: 45 Credit: 259,298 RAC: 0 |
HOLY CRAP!! I'm pretty sure these temps are NOT normal... Case: 104°F (40°C) CPU Diode: 168°F (75°C) CPU socket: 122°F (50°C) And that's without SETI running!! I don't understand, though. I've never had problems with this before, and everything looks clean... --- Computers Crunching: Laptop (1.4GHz, 512mb RAM) Main Laptop (Dual Core 2.16GHz, 2gb RAM) |
Digger Send message Joined: 4 Dec 99 Posts: 614 Credit: 21,053 RAC: 0 |
You're running an AMD, so i don't know how hot those can get... but i can tell you when i first started crunching SETI on my Intel Prescott, it heated up faster than my dad did when i broke his new fishing pole. I downloaded a popular temp monitoring and control program called Speedfan, which allowed me to see what was going on, and also tweak my fan speeds a little to keep my temps down. This has helped me quite a bit. An AMD person will be much better suited to tell you if those temps are critically high for a Barton. On a Prescott, the cpu would start throttling at those temps, but not shutdown. Sorry i can't be of any more help. P.S. You should also be able to check your temps from the system BIOS, which will allow you to doublecheck the accuracy of the readings you're getting from MBM. Dig |
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
> HOLY CRAP!! I'm pretty sure these temps are NOT normal... > > Case: 104°F (40°C) > CPU Diode: 168°F (75°C) > CPU socket: 122°F (50°C) > That's useless if you don't tell on which PC (you have 3 in your account) you measured these temps, when you read it and where (in the BIOS after a reboot? In Windows using what program?), what the outside environment was like at that moment (outside the case) and if you have sufficient cooling, case fans, CPU fan(s) etc on an adequate PSU. My CPU is pegged at 25C while my motherboard sensor says my mobo runs at 48C. If I turn off my PC for 10 minutes, then reboot and check my temps in the BIOS or in Windows in Aida32, my CPU is still pegged at 25C, the mobo will have gone down to 35C. Go figure. :) |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
> HOLY CRAP!! I'm pretty sure these temps are NOT normal... > > Case: 104°F (40°C) > CPU Diode: 168°F (75°C) > CPU socket: 122°F (50°C) > Caution: not all programs work well/accurately with all mother boards. 75 degrees C can cook food. and seems very high. Might wanna investigate the temp issue a bit further. Glad I could point you in a direction anyway. tony [edit] which directions are the case cooling fans blowing? I'd recommend the larger one(if it exists) blowing in, and the smaller one (if it exists) pushing the hot air out. any other suggestions people? [end edit] |
Scooley01 Send message Joined: 25 Apr 03 Posts: 45 Credit: 259,298 RAC: 0 |
My main computer is in my sig. I didn't think it needed to be reposted... --- Computers Crunching: Laptop (1.4GHz, 512mb RAM) Main Laptop (Dual Core 2.16GHz, 2gb RAM) |
Digger Send message Joined: 4 Dec 99 Posts: 614 Credit: 21,053 RAC: 0 |
My only suggestion right off the bat would be to first and foremost doublecheck those temps against what the BIOS is saying. As Tony stated MBM5 can give you some pretty wild readings until it is fine-tuned. You can't always rely on what it reports straight out of the box. The first and third readings (40°C/50°C) don't seem all that bad... lower than what my Prescott typically runs, but the 75°C reading seems a bit high. With that said however, i'm going to bow out and let the AMD experts help you. :) Dig |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
> With that said however, i'm going to bow out and let the AMD experts help you. > :) > > Dig > I'm bowing out also, I've never had a computer that allowed temp readings and I have no experience with MBM5, speedfan, etc. I'm just happy to narrow it down for you. good luck tony |
Chilean Send message Joined: 6 Apr 03 Posts: 498 Credit: 3,200,504 RAC: 0 |
|
Scooley01 Send message Joined: 25 Apr 03 Posts: 45 Credit: 259,298 RAC: 0 |
That's why I haven't posted SpeedFan results. It has given me overall lower temperatures, but there's still one at 75-76C HAven't checked BIOS yet, because that requires a restart, and i don't wanna restart LOL --- Computers Crunching: Laptop (1.4GHz, 512mb RAM) Main Laptop (Dual Core 2.16GHz, 2gb RAM) |
rajausa Send message Joined: 19 Feb 01 Posts: 25 Credit: 797,337 RAC: 0 |
It does sound like to me from what i have read that your computer is overclocked ; maybe not intentially but sounds like it . Look in bios and check your frequecies and FSB ; that may have changed on you without your intervention sometimes that happens . Your buss settings or memory timings may be too high for your cpu and or memory.If you are overclocked and know it; back off a little on your FSB.Hope that helps you out . |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 |
Every AMD based system I've built has required a slight upping of the CPU volts (5-10%) to get through MEMTEST86+ without errors. Memory test=CPU volts? Yes, memtest always gave me errors on the tests that have the CPU caches enabled. Apart from my XP3000/64 bit, all these processors ran windows and usual apps (Office etc) without needing extra volts. Another good stability test is PRIME95 (has stress tests built-in). As far as CPU temps are concerned, I find you cannot much rely on absolute values - there's plenty of discussions around about why. I look at the relative change from idle to 100% load. If a system is basically running Ok and the idle temp is say 58C and rises to 67C quickly under 100% load then back quickly to 58 when idle - then the CPU cooling is working IMO. (Quickly =10-15 seconds) |
DanielPharos Send message Joined: 12 May 03 Posts: 8 Credit: 58,255 RAC: 0 |
Disable 'Automatic reboot' if you haven't done that yet. 1. Right-click My Computer, and then click Properties. 2. Click the Advanced tab. 3. Under Startup and Recovery, click Settings to open the Startup and Recovery dialog box. 4. Clear the Automatically restart check box, and click OK the necessary number of times. 5. Restart your computer for the settings to take effect. Now, if Windows crashs, it will display a blue screen instead of rebooting. (If you see blue screens, google for the error message) Updating drivers sometimes helps... But according to those temps, it overheating. Got to your BIOS (update it if neccesary) and load the fail-safe defaults. |
Chilean Send message Joined: 6 Apr 03 Posts: 498 Credit: 3,200,504 RAC: 0 |
> It does sound like to me from what i have read that your computer is > overclocked ; maybe not intentially but sounds like it . Look in bios and > check your frequecies and FSB ; that may have changed on you without your > intervention sometimes that happens . Your buss settings or memory timings may > be too high for your cpu and or memory.If you are overclocked and know it; > back off a little on your FSB.Hope that helps you out . > That has happened with every PC that i have mal-overclocked.... they restart, and sometimes I fried the CPU...hehe... |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.