WU 803X

[hondo]

Hi hondo here from 51078 which is in somewhat of a decline now, just wondering if any of you guys are having any issues with WU # 803X series?

I've had plenty and no idea what the problem is hardware / software any pointers would be very much appreciated.

I'm folding -smps on the CPU with no problems whatsoever but frequently those coming off the GPU go through to 100% completion at which time the logfile reports an I/O error and just dumps the lot and starts again

I've run scandisk, an HD tools scan the drive is perfectly devoid of any defects, I know not now, what else can be done to cure the issue.

Thank for taking the time & for any help

Cheers hondo

 

[Biffa]

Not had any problems on my GTX470's with them

[14:50:27] Folding@home Core Shutdown: FINISHED_UNIT
[14:50:34] CoreStatus = 64 (100)
[14:50:34] Unit 4 finished with 98 percent of time to deadline remaining.
[14:50:34] Updated performance fraction: 0.980310
[14:50:34] Sending work to server
[14:50:34] Project: 8032 (Run 6, Clone 13, Gen 8)
[14:50:34] - Read packet limit of 540015616... Set to 524286976.


[14:50:34] + Attempting to send results [February 23 14:50:34 UTC]
[14:50:34] - Reading file work/wuresults_04.dat from core
[14:50:34]   (Read 406312 bytes from disk)
[14:50:34] Gpu type=3 species=20.
[14:50:34] Connecting to http://171.67.108.142:8080/
[14:50:39] Posted data.
[14:50:39] Initial: 0000; - Uploaded at ~79 kB/s
[14:50:39] - Averaged speed for that direction ~81 kB/s
[14:50:39] + Results successfully sent
[14:50:39] Thank you for your contribution to Folding@Home.
[14:50:39] + Number of Units Completed: 3258

 

[Ingleberry]

I had similar problem with the same series of work units, are you using Nvidia drivers 295.51?

If so revert back to 285.62 this solved the same type of error your reporting and also my games worked again too, the new drivers are not that good tbh widely reported on both folding@home and boinc sites.

 

[hondo]

Thanks Biffa, I wish I could say the same, I'm getting both hits & misses & am now at a loss as to why, all the misses are the same & each occurs only on the GPU. which is running @ 58c altho this can hit 64-65 depending on room temps

18:00:40] Completed  24750000 out of 25000000 steps (99%).
[18:02:54] Completed  25000000 out of 25000000 steps (100%).
[18:02:54] Finished fah_main status=0
[18:02:54] Successful run
[18:02:54] DynamicWrapper: Finished Work Unit: sleep=10000
[18:03:04] Reserved 356304 bytes for xtc file; Cosm status=0
[18:03:04] Allocated 356304 bytes for xtc file
[18:03:04] - Reading up to 356304 from "work/wudata_01.xtc": Read 356304
[18:03:04] Read 356304 bytes from xtc file; available packet space=786074160
[18:03:04] xtc file hash check passed.
[18:03:04] Reserved 44088 44088 786074160 bytes for arc file=<work/wudata_01.trr> Cosm status=0
[18:03:04] Allocated 44088 bytes for arc file
[18:03:04] - Reading up to 44088 from "work/wudata_01.trr": Read 44088
[18:03:04] Read 44088 bytes from arc file; available packet space=786030072
[18:03:04] trr file hash check passed.
[18:03:04] Allocated 544 bytes for edr file
[18:03:04] Read bedfile
[18:03:04] edr file hash check passed.
[18:03:04] Allocated 32556 bytes for logfile
[18:03:04] Read logfile
[18:03:04] GuardedRun: success in DynamicWrapper
[18:03:04] GuardedRun: done
[18:03:04] Run: GuardedRun completed.
[18:03:07] + Opened results file
[18:03:07] - Writing 434004 bytes of core data to disk...
[18:03:07] Done: 433492 -> 410218 (compressed to 94.6 percent)
[18:03:07]   ... Done.
[18:03:07] DeleteFrameFiles: successfully deleted file=work/wudata_01.ckp
[18:03:07] Shutting down core Problem in delete frame file
[18:03:07] Shutting down core 
[18:03:07] 
[COLOR=red][18:03:07] [/COLOR][COLOR=red]Folding@home[/COLOR][COLOR=red] Core Shutdown: FILE_IO_ERROR: Problem opening, closing or deleting a fil [/COLOR]
[18:03:09] CoreStatus = 75 (117)
[18:03:09] Error opening or reading from a file.
[18:03:09] Deleting current work unit & continuing...
[18:03:13] - Preparing to get new work unit...
[18:03:13] Cleaning up work directory
[18:03:13] + Attempting to get work packet
[18:03:13] Passkey found

 

[hondo]

I had similar problem with the same series of work units, are you using Nvidia drivers 295.51?

If so revert back to 285.62 this solved the same type of error your reporting and also my games worked again too, the new drivers are not that good tbh widely reported on both folding@home and boinc sites.

Thanks Ingleberry

He say totally exasperated "I honestly don't know any longer which drivers are in use whichever MS update recommended , however I'll certainly give 285.62 a try, things can only get better"

Thanks again

I'm off now to download & install 285.62 Cheers

 

[hondo]

285.62 wouldn't install, but I believe that I have just found out why.

Very old fashioned here, I'm still running Win XP pro SP3 285.62 is 64bit and doesn't like my OS.

I've un-installed whatever was on and downloaded & installed what Gigabyte recommends 263.09 that sound old enough for me. LMAO

I'm sorry if I sound / sounded rather exasperated, it's because ATM I am.

 

[Snapshot]

The latest working drivers for XP are 285.58 which can be downloaded from Nvidia's web site. 295.73 gave me a bad time as I reported here recently.

 

[hondo]

Good Morning Snapshot thanks I'll keep that in mind and will if required & give the driver you mention a try but I'd better stick with what I'm using now @ least until the WU I'm on now is completed,

BTW was it I/O errors?

Thanks

 

[Snapshot]

No, it was more that it just hung or reloaded the core. The thread's a bit further down the page with 'BEWARE' in the title.

 

[hondo]

Thanks very much Snapshot, I'll take a look at that now

BTW, I've not long since had another one go up the duff

 

[hondo]

No, it was more that it just hung or reloaded the core. The thread's a bit further down the page with 'BEWARE' in the title.

Yep, I now see what you mean, is it just me, or did this issue drive you nuts too?

OMG I've been bouncing off the ceiling here..

 

[hondo]

A month later, I'm using nvidia 296.10 now, & so far so good.