• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

File IO Error?

Tormond

[H]ard|DCer of the Month - June 2007
Joined
May 4, 2003
Messages
1,406
running the console version and the graphical version and teh console version just gave me a Folding@Home Core Shutdown: FILE_IO_ERROR. Any ideas? The graphical client is still pluuging along. Do I need to switch them both to the console version to make them play happy or do I need to change something in the console setup?
 
That would be my suggestion. Here is the guide with specifics on getting multiple consoles set up.

 
Ok I switched to 2 versions of console running as a service (running XP Pro btw) and will see how that works
 
Well after about 15 reboots I finally figured out why they wouldn't start as a service and all is good with the world. Is there any way to see what they are doing when they are running as a service (to make sure they haven't bombed out?)

 
Ok have EM3 up and running and my Dual 933 up and running as well as my X2. Guess I get to look for more boxes tomorrow :)
 
Tormond said:
Ok have EM3 up and running and my Dual 933 up and running as well as my X2. Guess I get to look for more boxes tomorrow :)

I hate to say it, but it looks like one (or more) of your machines is processing early end units,
hopefully it was just due to the graphical client.. but if you keep getting them its something
to look into..
 
Yea you've got early ends all over the place. Check the log file (\where-ever-you-instlled\FAHlog.txt) for the verbage EARLY_END I think is how they flag it.

Your stats @ EOC note the very low valued results returned. The lowest valued WU is ~40 points so anything less is not good.

Good luck!
 
ok it looks like the first core on my X2 is the one giving me the problem. I only have 1 early one on the 2nd core (and that was almost certainly me playing with it as it is near the top and has been running flawlessly since). All of the points listed in EM3 say at least 69 (the dual 933 has 2 of those) and a 241 (second core X2) and a 123 (first core X2). Now you have to understand that I have ZERO idea what any of this means :) but trying to give you everything that I can. What causes the Early End and how do I fix it? I can post teh log from that core if you need.

Edit: I uninstalled the graphical client before I installed the console client so all of these "errors" I guess you would call them have come from the console client

Trey

 
Well you're best bet now would to be Prime95 and Memtest the PC. Early ends are normally cause by issues in the CPU and/or the RAM. Since you're only getting early ends on the one core, it sounds like that could be the problem over the RAM since the other one is working fine.

 
Ok I will give that a shot. Hate to stop folding just to run Prime95 :)
 
Run around three loops of Memtest86 first.
Then run 8-24 hours of Prime95 if Memtest gave no errors.

Stanford expect the odd protien to fail as they push the perameters to the extream.
So I run about 1 in 100 Early Ends to good Work Units.
When I do a search of the FAHlox.txt file I look for "CoreStatus =".
Anything other than 64 is a bad end of a Work-Unit.

Luck........... :D
 
I ran memtest and Prime95 and they both were flawless. Figured out what the problem was however. I am running an Asus board and it was overclocking by default. Apparently one of my cores REALLY doesn't like being overclocked and folding at the same time (although it runs Prime95 etc with no problem) I set it to straight 200Mhz FSB (instead of the 208 or whatever Asus defaulted it to) and I haven't seen an Early End since

 
Tormond said:
I ran memtest and Prime95 and they both were flawless. Figured out what the problem was however. I am running an Asus board and it was overclocking by default. Apparently one of my cores REALLY doesn't like being overclocked and folding at the same time (although it runs Prime95 etc with no problem) I set it to straight 200Mhz FSB (instead of the 208 or whatever Asus defaulted it to) and I haven't seen an Early End since


That's actually somewhat weird. Usually, if memtest and Prime95 don't have any problems, the system is usually about as stable as you can get as far as CPU and RAM are concerned. If a system passes those tests, F@H will usually not have a problem. Anyway, as long as it's stable it's all good now.

 
Is Prime95 multithreaded though? I don't use it, and if it's only using one of the cores, that could explain something. I've seen this auto setting for overclocking my Asus board, but I've never had it automatcially set an OC.

 
Prime is only single threaded.
I need to run two copies to fully stress my duallies.

Luck......... :D
 
You know. I still havent run prime95 on anything i own yet.

where do you d/l it from anyways?
 
not my friend. although it was founded by fellow stanford people :D
- Fold [H]ard - - Or go [H]ome -
 
I was assuming that since it was my first core that was crapping out that Prime would use that one (as most things do). It could very well have been a specific set of cores that I was getting. I did delete the whole folder and reset it up from scratch so that may have been the problem as well

 
I agree. I just wish that the stats would update as I have finished 2 241 pointers today :)
 
Tormond said:
I agree. I just wish that the stats would update as I have finished 2 241 pointers today :)

Ahhh nothing like the feeling of having points in the bank and then waiting for the stats to update. I was like that the other morning knowing that I had hit my 200k, but my badge wasn't showing it.

 
Tormond said:
I was assuming that since it was my first core that was crapping out that Prime would use that one (as most things do).
Windows automatically migrates processes from one processor to another. Thus, it could have been running on either one. If you manually modify the affinity, then it'll run on whatever you set it to, but otherwise it goes for whichever is least loaded.

 
One thing people don’t think much about is corrupted folding files. That happens far more then you realize and the fix is simple. Replace the main folding .EXE file and delete all existing core files. Restart the program and it will grab a new core.

The core files and the EXE file are a bit sensitive to being shut down improperly or forced to fail in an unstable CPU situation.

Give it a shot, don’t cost nuttin;)


 
BillR said:
One thing people don’t think much about is corrupted folding files. That happens far more then you realize and the fix is simple. Replace the main folding .EXE file and delete all existing core files. Restart the program and it will grab a new core.

The core files and the EXE file are a bit sensitive to being shut down improperly or forced to fail in an unstable CPU situation.

Give it a shot, don’t cost nuttin;)



I have having this problem a lot with a borg I had. No matter what it seemed to crash a lot. The system, or at least the F@H client was being shut down improperly a lot and even when it wasn't, there were a lot of client/core communication problems. I deleted the core, a new one was downloaded and haven't had hardly any trouble since.

 
I guess I'll just consider myself lucky then. The entire time I've folded, I've passed the same set of executeables around my home boxen and haven't had a problem. So would it be safe to assume that maybe be the first suggested fix before freaking out and running Prime and Memtest?

 
Well it certainly seems that way to me now after wasting some very needed cycles on Prime and Memtest

 
Back
Top