6.34 Ubuntu 10.04 Terminal Crashes

[Bijiont]

I know there aren't many native *nix folders however I am hoping someone will have an idea or two.

I have 6.34 client installed on 10.04 Ubuntu using the Stanford instructions. client loads just fine with no errors however after about 1%~2% progress the terminal window and FAH along with it will crash. Now the system runs just fine and doesn't hang as I am able to reload it right away.

The log also doesn't indicate any errors from what I am able to see.

# Linux SMP Console Edition ###################################################
###############################################################################

                       Folding@Home Client Version 6.34

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: /home/stealthdragon/folding
Executable: ./fah6
Arguments: -smp -verbosity 9 -bigadv 

[11:14:58] - Ask before connecting: No
[11:14:58] - User name: Bijiont (Team 32)
[11:14:58] - User ID: 
[11:14:58] - Machine ID: 1
[11:14:58] 
[11:14:58] Loaded queue successfully.
[11:14:58] 
[11:14:58] + Processing work unit
[11:14:58] Core required: FahCore_a5.exe
[11:14:58] Core found.
[11:14:58] - Autosending finished units... [April 15 11:14:58 UTC]
[11:14:58] Trying to send all finished work units
[11:14:58] + No unsent completed units remaining.
[11:14:58] - Autosend completed
[11:14:58] Working on queue slot 01 [April 15 11:14:58 UTC]
[11:14:58] + Working ...
[11:14:58] - Calling './FahCore_a5.exe -dir work/ -nice 19 -suffix 01 -np 8 -nocpulock -checkpoint 15 -verbose -lifeline 1590 -version 634'

[11:14:59] 
[11:14:59] *------------------------------*
[11:14:59] Folding@Home Gromacs SMP Core
[11:14:59] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[11:14:59] 
[11:14:59] Preparing to commence simulation
[11:14:59] - Ensuring status. Please wait.
[11:15:08] - Looking at optimizations...
[11:15:08] - Working with standard loops on this execution.
[11:15:08] Examination of work files indicates 8 consecutive improper terminations of core.
[11:15:11] - Expanded 25461284 -> 31941441 (decompressed 125.4 percent)
[11:15:11] Called DecompressByteArray: compressed_data_size=25461284 data_size=31941441, decompressed_data_size=31941441 diff=0
[11:15:11] - Digital signature verified
[11:15:11] 
[11:15:11] Project: 2686 (Run 1, Clone 13, Gen 88)
[11:15:11] 
[11:15:11] Entering M.D.
[11:15:17] Using Gromacs checkpoints
[11:15:18] Mapping NT from 8 to 8 
[11:15:23] Resuming from checkpoint
[11:15:24] Verified work/wudata_01.log
[11:15:24] Verified work/wudata_01.trr
[11:15:24] Verified work/wudata_01.xtc
[11:15:24] Verified work/wudata_01.edr
[11:15:24] Completed 20498 out of 250000 steps  (8%)

I have also been checking the system logs however have come up dry there as well. I have considered trying Ubuntu 10.10 however would prefer the easy fix over redoing the entire system.

Thanks in advance and outside of going out for a couple hours I will be here working on this until I can get it working correctly.

 

[ChasR]

My guess is FAH is crashing and the terminal is closing as a result. Stability would be the first thing to look at.

It is writing checkpoints so it's not likely to be disk related, but are you losing the WU occasionally? That may point back to an intermittent disk problem. Writes a checkpoint and fails to write to the disk or log on the next checkpoint.

Any chance its swapping on the checkpoint write, where memory use would peak?

One stick of ram not seated properly?

I'll think on it some more, but having a number of 10.04 Ubuntu installs without ever experiencing what you describe, leads me to believe it is a hardware issue rather than a Linux issue.

 

[Bijiont]

My guess is FAH is crashing and the terminal is closing as a result. Stability would be the first thing to look at.

It is writing checkpoints so it's not likely to be disk related, but are you losing the WU occasionally? That may point back to an intermittent disk problem. Writes a checkpoint and fails to write to the disk or log on the next checkpoint.

Any chance its swapping on the checkpoint write, where memory use would peak?

One stick of ram not seated properly?

I'll think on it some more, but having a number of 10.04 Ubuntu installs without ever experiencing what you describe, leads me to believe it is a hardware issue rather than a Linux issue.

After posting I dropped the turbo down to 4.4ghz and so far so good. This is the first WU after the new install. All the hardware was checked prior to install.

As of right now it appears to have been stability like you said however I was expected the machine top hang rather than just FAH. Will keep and eye on it for the next few hours and see what happens.