Stable system (months) suddenly becomes unstable?

A

Aix.

[H]ard|Gawd
Joined
Sep 30, 2010
Messages
1,959
I have two systems: a 980X BigAdv rig and an SR-2 dual-L5640 BigAdv rig.

A couple of months ago my 980X rig - which had been stable for weeks - choked on a P2686 WU and started hanging on certain frames of the WU. It turned out that bumping my vCore up a bit fixed the problem and the rig continued folding without issue afterward. Hooray.

Last night I came home to find my SR-2 rig - which had been stable for 3+ months - had restarted for whatever reason. I loaded the FAH tracker back up and noted that it was crunching a P2686 WU. After a few hours the client hung in the very same way that my 980X had hung in the past (stuck on one frame forever, even though all cores were at 100%). I stopped the client, tried bumping voltage in my BIOS from 1.35 to 1.365~ and restarted the tracker...only to find that my progress had reset to 0%.

Has anyone else had something like this happen? The only other issue I had with this system was when I had a stick of RAM fail, but all 12GB are reporting this time. Other than that all my settings have worked from day one without issue. Should I be looking at something else than the CPUs?
 
shit happens.


yes ive had this issue but no idea what causes it. havent had it in a long time now.
 
yeah it happens. especially if you're right at the edge of stable and unstable. as far the 0% thing goes. it probably wasn't able to off load the ram data(WU data) to the hard drive which caused the WU to restart.
 
still cant figure out why my one machine restarts itself when a GPU wu gets to 100%.
 
Aix. said:
Last night I came home to find my SR-2 rig - which had been stable for 3+ months - had restarted for whatever reason. I loaded the FAH tracker back up and noted that it was crunching a P2686 WU. After a few hours the client hung in the very same way that my 980X had hung in the past (stuck on one frame forever, even though all cores were at 100%). I stopped the client, tried bumping voltage in my BIOS from 1.35 to 1.365~ and restarted the tracker...only to find that my progress had reset to 0%.

Has anyone else had something like this happen? The only other issue I had with this system was when I had a stick of RAM fail, but all 12GB are reporting this time. Other than that all my settings have worked from day one without issue. Should I be looking at something else than the CPUs?
Click to expand...

I have one SR2 rig that suffers from long frame times / hung frames - particularly on days of high humidity - go figure.

This rig is otherwise folding stable for weeks, passes IBT etc. But it needs 2 notches more vcore to reliably solve the long frame issue. PITA.
 
Aix. said:
Has anyone else had something like this happen?
Click to expand...
Yes, one system in my farm is plagued by this problem and it only happens with -bigadv WUs, it doesn't matter which project. Note, this system has been stable for a very long time; since -bigadv was released more than a year ago. Whenever I switch to regular SMP, all is well. Lowering the frequency to stock does absolutely nothing and no other software seems affected. This to me would normally point towards bad WUs but only one system is plagued by restarts and it has been a recent development.

Should I be looking at something else than the CPUs?
Click to expand...
If it is indeed a hardware problem (I have my doubts) I would first look at memory modules then the motherboard. I already swapped modules and it had no noticeable effect. Processors are resilient unless you run very high voltages with poor cooling for long periods. In my case, the motherboard does not support voltage adjustment and my OCing is quite limited as a result. My CPUs aren't the problem, a supposition supported by having the reboot issue occur even at bone stock speeds.

Tobit did mention recently that there are a bad crop of WUs in the wild. If this is the problem, it seem strange that only some systems are susceptible to failure from a corrupt batch. Personally, it doesn't annoy me all that much because my days in this project are numbered, but I could understand the frustration some people must be having with this issue.
 
Filter said:
when and why are you quitting apollo?
Click to expand...
I won't get into the why but I'm leaving when I hit the 100M milestone or when the weather turns warm enough that the heat from my small farm is no longer a benefit, whichever arrives second.
 
APOLLO said:
I won't get into the why but I'm leaving when I hit the 100M milestone or when the weather turns warm enough that the heat from my small farm is no longer a benefit, whichever arrives second.
Click to expand...

It will be a sad day Apollo when you shutdown.:(:(
4th in the [H]orde and no. 52 in the project is a stunning achievement.:cool:
 
Yes, you would be missed, APOLLO. As long as I have been here you have always been right there at the top. Whatever the reason, best wishes and you can always change your mind and come back.
 
APOLLO said:
I won't get into the why but I'm leaving when I hit the 100M milestone or when the weather turns warm enough that the heat from my small farm is no longer a benefit, whichever arrives second.
Click to expand...

Oh no... :(

I hope you decide to change your mind about this, as we've all had some really good times folding for the team.
 
Sounds like dodgy WU, delete it and try another. Hopefully it isn't hardware starting to fail.

APOLLO said:
I won't get into the why but I'm leaving when I hit the 100M milestone or when the weather turns warm enough that the heat from my small farm is no longer a benefit, whichever arrives second.
Click to expand...

It must be something in the atmosphere up there Flecom left just after 100M too. It is a shame non the less. :(
 
Wiseguy2001 said:
It must be something in the atmosphere up there Flecom left just after 100M too. It is a shame non the less. :(
Click to expand...

Flecom has had to shut down due to him losing his job and no longer being able to afford to fold 300,000 points per day.:( Hopefully once he is back on his feet he will be back:).

Apollo has given some insight into his reasons for his impending shut down but they are scattered comments across a wide range of posts and not easily listable.:(
 
Wiseguy2001 said:
That's a shame, I remember from back along that he has skills. He'll land on his feet.
Click to expand...

Hope so, he only came back to the team in force in april/may last year. Its a shame to lose hime again but fully understand why
 
Wiseguy2001 said:
Sounds like dodgy WU, delete it and try another. Hopefully it isn't hardware starting to fail.
Click to expand...

Yup, I just finished a 6900 without issue at the same settings. Hopefully I'm good for a while again!
 
I'm glad it isn't just me! Mine started restarting for no apparent reason, screen would go black and I'd hear the HDD click. It folded all weekend long and then today it started rebooting a lot! I would start typing and bam, it was rebooting again! I thought I had killed my video card, but then after she rebooted I thought maybe it was the CPU dying since it rebooted while I was lowering my clocks in the BIOS!

After all of that it wouldn't even boot at all. I was sad and thought that it was a bad time for Sandy Bridge to have a hardware errata that led to all boards being pulled. But I let her sit for half an hour and she booted again, haven't fired up the clients yet because I really can't have my PC die on me right now :( When I get my backup PC a video card I may fire up the clients again so long as I can order new PC parts if this one bites the dust!
 
You must log in or register to reply here.
Back
Top