600point WU restarted from scratch 3 times?!?

N

neisius

Gawd
Joined
Mar 11, 2004
Messages
783
"[22:09:06] Preparing to commence simulation
[22:09:06] - Ensuring status. Please wait.
[22:09:24] - Assembly optimizations manually forced on.
[22:09:24] - Not checking prior termination.
[22:09:55] - Expanded 2970644 -> 16166417 (decompressed 544.2 percent)
[22:09:58] - Checksums don't match (work/wudata_05CP.arc)
[22:09:58] - Starting from initial work packet"

So I get ~20hours into this 75hour WU and reboot the system and the WU starts from scratch. This has happened at least 2-3times in the last few days with the same WU. I guess when I restart my system something goes wrong and it doesn't save where it left off.
Any ideas how to stop this from happening?

 
Yes! This happens when the box gets rebooted/shut down and F@H doesn't get a chance to save the progress. The easiest way I know to prevent this is to stop the service manually before the restart! ;)
Hope this helps!


Thanks xENo!!!
 
are you absolutely sure it is the same wu, it might be the same protein but a different wu.

The reason I am saying this is because in one of my oc attempts, the memory was taken too far and would dump the wu when XP started and download a new one from Stanford. So if you are on an OC, drop the memory a bit.
 
Same thing happen to me last nite. I rebooted and it started from 0 but my other instance which is also working on a 600 pointer started back up fine.
 
magnusvir said:
So if you are on an OC, drop the memory a bit.
Click to expand...
everything is running at stock speeds(not my choice)...except for the vidcard.

I think the frist time it happened was when I tried going to xbox-linux.org and explorer crashed :D

 
try running it without the forceasm flag on. I have had problems sometimes with it running. Just monitor and see it starts with SSE boost.
 
If your box quits in the middle of something. F@H will not be able to save a check point.
Sometimes if the box shuts down fast enough it will not save a checkpoint!
Checksums do not match......starting from beginning!


Thanks xENo!!!
 
If the issue persists you really might want to think about not folding the big units. They do tend to stress your machine a bit more and if your prone to rebooting a lot your not getting any "work" done so to speak. Make the switch back to the regular units and your production will probably go up.

Just a thought

Luck
 
magnusvir said:
try running it without the forceasm flag on. I have had problems sometimes with it running. Just monitor and see it starts with SSE boost.
Click to expand...
I think I'll turn off the -forceasm and -advmethods for awhile.

Yeh this system gets alot of use and is only on ~15hr/day. I'll leave my other system to do tthe bigger units.

 
What client are you using and is it running as a service?

I have had no problems running it as a service on any machine as long as it's shut down normally. I've had a couple of problems with the program just running normal in the taskbar when I shut down normally and didn't shut the program down first. Ever since I started shutting it down manually before I restart the system, I have had no problems.

 
SmokeRngs said:
What client are you using and is it running as a service?

I have had no problems running it as a service on any machine as long as it's shut down normally. I've had a couple of problems with the program just running normal in the taskbar when I shut down normally and didn't shut the program down first. Ever since I started shutting it down manually before I restart the system, I have had no problems.

Click to expand...
I'm running the GUI one not as a service. maybe the console one as a service would work better for me.

 
A lot of people have problems with the graphical client. I have always used the console client myself since I don't really need to see the proteins as they process and if I really want to, I can use EMIII for that. I have never had a problem with the console version other than what I can attribute to myself or my system.

 
SmokeRngs said:
A lot of people have problems with the graphical client. I have always used the console client myself since I don't really need to see the proteins as they process and if I really want to, I can use EMIII for that. I have never had a problem with the console version other than what I can attribute to myself or my system.

Click to expand...
looks like the console client is a whole 30sec/frame faster too :eek: :D

 
You must log in or register to reply here.
Back
Top