Having crash problem with SETI <solved>

[NeoPTLD]

http://forums.anandtech.com/me...2&amp;threadid=1406793

I'm having random blue screen problem when I have SETI going. I ran memtest86 and I got no error in 14 hours. I didn't have this problem last year and I'm using the same SETI client (CLI w/ SETI driver)

I added RAM since last year, but if it passes memtest, it couldn't be the problem could it?

help
memtest86 result Could it still be memory?

or is it the CPU?

CPU chipped ! !

 

[Wiz]

Are you overclocking?

(also replied in your other thread)

 

[keyboardcaper]

it could be a few things, you can try www.ultimatebootcd.com they have a few progs to test your memory, i recommend goldmemory.cz, i would say its your power supply, if your motherboard has it check your bios for a hardware monitor, see your voltages, if they differ by more than (+-5% would be acceptable to replace your power supply, TeAm what do you think yes? no?) or if you dont wanna reboot, checkout motherboard monitor to check your voltages, heat doesnt seem to be your issue, but i bet its your power supply

 

[Orange Kid]

could just be the client has gone flaky, try removing and reinstalling 🙂

 

[NeoPTLD]

Last year:
Windows XP Pro
512MB memory
Ran SETI all night long, used my computer while SETI ran and no crashes ever.

This year:
I installed SP2

added another 512MB of memory to make 1GB total, but it passes memtest86 just fine with the same memory settings as last year. If memory was the issue, wouldn't memtest86 complain before SETI would?

I'm using the same power supply from last year.

No overclocking anywhere, ever!

When I leave the computer running with SETI, it will eventually BSOD(which reboots the computer after it dumps the physical memory to disk) then I find it at login windows. If I browse the web with SETI CLI running in idle mode, Mozilla have ocassional crashes.

None of these happens if SETI is not running.

 

[lobadobadingdong]

right click on my computer
select properties
select the advanced tab
select settings under the Startup and Recover box
uncheck the box to Automatically restart under the System failure box
hit OK
OK again
and restart your computer.

then run your machine until it crashes, it won't automatically restart on you now, so that you can get the stop code to see what the actual problem is. It could be any number of things if its the seti cli, it will actually give you something like this.

STOP 0x00000000 (0x00000000, etc.) fault in module setiathome-3.03.i356-winnt-cmdline.exe

report what the first two 0x00000000 codes are and I can look it up on Microsoft's technet to find out exactly what is causing the fatal error.

It's possible that your video card's cooling is no longer doing the job well enough when under a long term strain on your system. or any number of other issues, including failed active-x modules, invalid shortcuts in the registry, etc.

 

[NeoPTLD]

I will go ahead and do what you said. Do you know if Windows keeps a log of the last few errors? I'd like to compare if all the blue screen incidents came from the same error.

 

[lobadobadingdong]

occasionally you can find information in the event viewer (under the system tab), but many times the data is mis-leading to the actual problem.

you can find event viewer in the control panel, in adminstrative tools.

 

[NeoPTLD]

My friend was suggesting it was overheat. Temperature looked fine, but I removed the HS + CPU for an inspection and found a small core damage. It may have chipped when the computer got bumped a bit on a hand truck.

Heatsink itself weighs somewhat over a pound and this probably doesn't help anything.

cpubroken.jpg

Do you guys suppose its the CPU?

 

[Assimilator1]

Dunno ,does it pass Prime95 torture tests?

 

[NeoPTLD]

Originally posted by: lobadobadingdong
right click on my computer
select properties
select the advanced tab
select settings under the Startup and Recover box
uncheck the box to Automatically restart under the System failure box
hit OK
OK again
and restart your computer.

then run your machine until it crashes, it won't automatically restart on you now, so that you can get the stop code to see what the actual problem is. It could be any number of things if its the seti cli, it will actually give you something like this.

STOP 0x00000000 (0x00000000, etc.) fault in module setiathome-3.03.i356-winnt-cmdline.exe

report what the first two 0x00000000 codes are and I can look it up on Microsoft's technet to find out exactly what is causing the fatal error.

It's possible that your video card's cooling is no longer doing the job well enough when under a long term strain on your system. or any number of other issues, including failed active-x modules, invalid shortcuts in the registry, etc.

What happens is that things like Mozilla, Windows explorer and such and such crash when SETI is running. These programs don't crash otherwise.

I cleared the error log and let SETI run overnight. Unfortunately, I forgot to set the computer to not autoreboot, but this is what I pulled from the log.

Event Type: Error
Event Source: System Error
Event Category: (102)

Description:
Error code 0000000a, parameter1 00000010, parameter2 00000002, parameter3 00000001, parameter4 804f6c7b.

UGH... crashed again:

BSOD screen shot (greyscaled to save server bandwidth)

 

[Grumpy1]

I had a similar proablem with two of my machines. Both machines would reboot at the same time in the SETI process. Just after sending a WU and starting the next, after the first report of estimated time both machines would reboot. Microsoft most of the time would report a driver proablem usually pointing at the video driver.

One machine an unlocked XP2500 running at 3000 with a 1024 stick of infinion 2100 memory. I suspected the memory but memtest indicated it was OK even at fast memory timings. The proablem was solved by bumping the CPU voltage one notch.

The other machine a Palomino 2100 running a 512 stick of Staples 2700 crap. Bumping the voltage did no good, slowing down the memory timings solved the proablem.

Stick with it you will find the right combination.


The Grump

 

[Assimilator1]

Neo

Try running the Prime95 test ,you may have a hardware fault

 

[lobadobadingdong]

everything I found on technet for that (Page File Number List) points to either a dirver, or a memory problem.

You might try alternating using only one schtick or the other of your memory (in different slots) to see if the problem changes or goes away, if doesnt' go away, I'd make a backup and re-load windows (and your drivers from scratch)

 

[NeoPTLD]

It was memory settings. In my BIOS, there's an option called "Top Performance". It was enabled, so I turned it off and blue screen stopped.

Before: CAS 2.0, RAS to CAS delay 2, RAS precharge 2, cycle time 7, bank cycle time 5 DRAM idle timer 8
Now:CAS 2.0, RAS to CAS delay 3, RAS precharge 3, cycle time 7, bank cycle time 10 DRAM idle timer 8

One thing I don't understand is memtest86 never reported any error in the previous settings after 48 hrs of intensive testing.

SETI ran fine in previous setting with either stick of RAM. Problem only happens when I have both in at the same time in previous settings. With new settings, my computer seems to be stable even with SETI running.

 

[Wiz]

Problem only happens when I have both in at the same time

Yes, I've found this to be true in my own life experience...

Using one stick I can just about always run faster than using two sticks in any of my systems.
The tradeoff is often having more ram, which apps love, for having slower mem timings.
The answer? 1GB sticks of course - but then you will end up wanting more than one and be back in the same spot 🙂

 

[Assimilator1]

Yeah makes sense ,more modules means more problems with keeping 'clean signals' on the mbrd.
Anandtech did this article recently about memory timings which explain it.

My 2nd rig was running at 150 MHz FSB/memory with Crucial PC133 CL2 ,with 1 256MB stick it was fine at that ,but when I added a 128MB stick it became very unstable so I had to drop it down a step to 145MHz

Glad you've got it sorted anyway🙂
Btw it would seem memtest isn't very good! (I'm sure Prime95 would of picked it up😛,which is why I suggested it😉)

 

[Smoke]

When you replied to Wiz, "No overclocking anywhere, ever!", I was thrown off the track.

Overclocking is tweaking your mem settings as well as your FSB.

Really glad to see you have figured out the reason for the BSOD.

In my experience, OCing to get faster Seti times is a lot of fun but it really does hurt your overall production. The slight increase in speed usually never makes up for the ground lost during the down times.

Thanks for helping the TeAm ... crunch on. :beer:😀:beer:

 

[Assimilator1]

The idea is that you sort out any overclocking problems in the 1st few days of trying it😉 ,Prime95 is an ideal tool for doing that.Oh &amp; their was nothing small about the increase I got out of my old XP1700 ,1.47GHz > 2.04GHz ,about 38% 😀 ,over about 2.5yrs that's a lot more WUs when its running 24/7🙂.Its fairly common to get that kind of overclock out a of a low to midrange CPU
Its only really when people try to get the last few % out of an already high clocked/overclocked CPU when you loose more than you gain because of crashes.

 

[dmcowen674]

Originally posted by: NeoPTLD
It was memory settings. In my BIOS, there's an option called "Top Performance". It was enabled, so I turned it off and blue screen stopped.

Before: CAS 2.0, RAS to CAS delay 2, RAS precharge 2, cycle time 7, bank cycle time 5 DRAM idle timer 8
Now:CAS 2.0, RAS to CAS delay 3, RAS precharge 3, cycle time 7, bank cycle time 10 DRAM idle timer 8

One thing I don't understand is memtest86 never reported any error in the previous settings after 48 hrs of intensive testing.

SETI ran fine in previous setting with either stick of RAM. Problem only happens when I have both in at the same time in previous settings. With new settings, my computer seems to be stable even with SETI running.

Cool 😎 Isn't that something how the DC program showed the weakness in the system?

Another use of SETI 😀 :thumbsup: