F@H and Freezing Computers

[stratman]

I got up this morning, hoping to find my my third WU completed. It was a big one -- I had spent 3+ days on it.

I turn my monitor on, fire up the F@H log file, and am treated with this sight:

[03:43:35] Preparing to commence simulation
[03:43:35] - Ensuring status. Please wait.
[03:43:51] - Looking at optimizations...
[03:43:51] - Working with standard loops on this execution.
[03:43:51] - Previous termination of core was improper.
[03:43:51] - Files status OK
[03:43:52] - Expanded 352404 -> 1770005 (decompressed 502.2 percent)
[03:43:53] - Checksums don't match (work/wudata_03.xtc)
[03:43:53] - Starting from initial work packet

OMG. COMPLETING "p1409_polyQ26 in water" would have got me like SEVERAL milestones.

My question is this: does having your comp. freeze always cause F@H to start the WU over?

Because my computer (1ghz athlon 256 ram) freezes quite often...

 

[3chordcharlie]

It can cause that.

Especially if there's any sort of instability, it can interfere with proper writing to the disk.

I have a palomino 1.2ghz that had this problem constantly; sometimes even on a normal reboot. It wans't OC'd but I finally just upped the cpu voltage one notch, and it hasn't happened since (I guess my motherboard didn't supply enough voltage).

 

[ProviaFan]

Any instability can cause F@H to get screwed up and terminate WUs early, or restart WUs. The suggestion of increasing the voltage slightly (a _very tiny_ amount) is good, but check your cooling first to make sure that filters aren't clogged with dust, fans still work properly, etc.

 

[stratman]

K thanks guys

I'll check out my bios options.

 

[GLeeM]

stratman, sorry about losing all those hours. Same thing has happened to me.

Couple things might help:

A journaling file system. (NTFS or that new Linux fs)

If you have FAT/FAT32 filesystem, look for a way to skip checkdisk on startup.

Use the -forceasm flag so that you don't see: "Working with standard loops on this execution." after a restart.

Thanks