Odd Error 2nd time

[WarriorII]

[04:10:29] Completed 275000 out of 500000 steps (55 percent)
[04:22:55] Writing local files
[04:22:55] Completed 280000 out of 500000 steps (56 percent)
[04:35:21] Writing local files
[04:35:21] Completed 285000 out of 500000 steps (57 percent)
[04:45:37] - Autosending finished units...
[04:45:37] Trying to send all finished work units
[04:45:37] + No unsent completed units remaining.
[04:45:37] - Autosend completed
[04:47:46] Writing local files
[04:47:46] Completed 290000 out of 500000 steps (58 percent)
[04:47:46] - Checksums don't match (work/wudata_01.xtc)
[04:47:47] Premature end of file when checksumming (11886692 bytes left)
[04:47:47] - Could not calculate checksum (work/wudata_01.xtc)
[04:47:49] Checksum not what expected.
[04:47:51]
[04:47:51] Folding@home Core Shutdown: FILE_IO_ERROR
[cli_1]: aborting job:
Fatal error in MPI_Wait: Error message texts are not available

This is the 2nd time this happened.
1st time was when I was away this past week & I have no idea what happened.

I deleted the files & restarted the client. (native linux/SMP)

1st time, I thought maybe it was the oc. So I reduced it & re-checked the voltages. bumped up the vcore to 1.42 from 1.40. mem @ 2.23
(2.2 rated TeamX )

Now it happened again.
Anyone else having a problem?

I haven't had any type of problem on these B4.
My other linux smp box is running them just fine.

Maybe I'll chek with memtest86+ & see how it goes.

:shrug:

[harlam357]

Humm, could it be a failing HD? Anyone know how to do a file system check in Linux? :shrug:

The only time I've gotten a FILE_IO_ERROR is in situations where one client got in the way on another. :rolleyes: Which was my stupidity in the first place. ;)

Those experiences were with standard clients... I haven't seen such an error on SMP yet. ;)

[pscout]

fsck is the file system check command.

I have run it directly myself after linux got corrupted (hdd failing) , and ubuntu seems to do it automatically every 25 boots or so.

[benbaked]

Humm, could it be a failing HD? Anyone know how to do a file system check in Linux? :shrug:


use fsck at the command line.

doh, pscout beat me to it. My buddy had a similar error to this on his e4300/ecs combo. Apparently he "resolved" it the same, by raising the vcore and vdimm slightly, and reinstalling the linux smp client. I wonder if it will return on his machine? He's running suse 10.2.

[leelegend]

have you stopped and started the core?, i get this quite a bit on the macs

lee

[WarriorII]

Yes, I have restarted the core.
Reduced the oc on it & so far it seems fine.

Makes me wonder about the memory. :-/

[WarriorII]

Problems, problems, problems.

Ran MemTest86+ ver 3.3 (newest) and right from the git-go, errors.

Tests 2,3,4 right off the start. I didn't bother letting it finish.

Ran it @ 350mhz, 1:1, DDR2 701
Ran it @ 266mhz 1:1, DDR2 800

It's toast. :(

Replaced it with Crucial DDR2 8500 sticks. (1066ers)
Runs fine now. (go figure) :p

Moral of the story, don't skimp on good memory.
If buying used, ask for a run on Memtest to be sure.