QMD core shutting down

DoctorGeo2008DoctorGeo2008 Flint, MI
edited July 2005 in Folding@Home
Ok, I did some searching through Stanford's forums and SM's forums and I wasn't able to find an answer to my question. I had the -advmethods set on 1 of the 2 consoles of FAH that I'm running on my work duel Xeon computer. Here is a copy of what I got EVERY time I ran the QMD core:

[23:02:06] + Processing work unit
[23:02:06] Core required: FahCore_96.exe
[23:02:06] Core found.
[23:02:06] Working on Unit 00 [July 11 23:02:06]
[23:02:06] + Working ...
[23:02:06]
[23:02:06] *
*
[23:02:06] Folding@Home QMD Core
[23:02:07] Version 1.04 (Apr 7, 2005)
[23:02:07]
[23:02:07] Preparing to commence simulation
[23:02:07] - Looking at optimizations...
[23:02:07] - Created dyn
[23:02:07] - Files status OK
[23:02:07] - Expanded 260832 -> 826210 (decompressed 316.7 percent)
[23:02:07]
[23:02:07] Project: 1912 (Run 321, Clone 9, Gen 15)
[23:02:07]
[23:02:07] Writing local files
[23:02:08] Extra SSE2 boost OK.
[23:02:08] Entering QMD...
[23:02:43] System: p1912_ALA-dipeptide_umb2
[23:02:43]
[23:02:43] Performing initial WF calculations
[23:02:43] - Number of total steps will change until convergence
[23:04:04] Completed 0 out of 2000 steps (0)
[23:17:15] Timered checkpoint triggered.
[23:17:19]
[23:17:19] Folding@home Core Shutdown: FILE_IO_ERROR
[23:17:23] CoreStatus = 75 (117)
[23:17:23] Error opening or reading from a file.
[23:17:23] Deleting current work unit & continuing...
[23:18:28] - Preparing to get new work unit...
[23:18:28] + Attempting to get work packet
[23:18:28] - Connecting to assignment server
[23:18:28] - Successful: assigned to (171.64.122.113).
[23:18:28] + News From Folding@Home: Welcome to Folding@Home
[23:18:28] Loaded queue successfully.
[23:18:30] + Closed connections
[23:18:35]

It appears that it does the simulation in about 13+ minutes and when it tries to do a timer triggered checkpoint it can't read the info. Some mentioned that it could be a hard drive permissions issue but I haven't had any problems to date with other cores. Some mentioned that I may be running out of HD space. This may be possible as I don't know what my quota is on this network drive. So...my questions are....


1) Should I move FAH to the C: drive since I'm currently running it off of a network drive? I would then know for sure that I have enough HD space.

2) Could the QMD's be faulting out because Hyper Threading appears to NOT be enabled on my computer?

Thanks,

Geoff

Comments

  • mmonninmmonnin Centreville, VA
    edited July 2005
    Some people have had problems runing 2 GMD cores at once.
  • DoctorGeo2008DoctorGeo2008 Flint, MI
    edited July 2005
    mmonnin wrote:
    Some people have had problems runing 2 GMD cores at once.

    Do you mean QMD instead of GMD?? If so, I'm only running one QMD. The other is a Gromacs.
  • mmonninmmonnin Centreville, VA
    edited July 2005
    Yeah I meant QMD.
  • DoctorGeo2008DoctorGeo2008 Flint, MI
    edited July 2005
    I think I figured it out... It appears that our network was down during the time that the QMD was running. And, the work file itself is about 175 Megs and I don't believe that our quota allows that much space to be taken up. So... I did what I should have done the FIRST time and I'm running them from the C: drive. Seems to be working fine now.
  • profdlpprofdlp The Holy City Of Westlake, Ohio
    edited July 2005
    Wow - that's a hefty WU. :wow2:

    Glad you got it figured out. :)
  • DoctorGeo2008DoctorGeo2008 Flint, MI
    edited July 2005
    Must be the size varies.. it's only 56 megs right now but it's using 230 Megs of RAM! Good thing these work machines have 3 GIGS of RAM!!! :thumbsup:
  • profdlpprofdlp The Holy City Of Westlake, Ohio
    edited July 2005
    I don't know if I'm more amazed by the fact that it's using 230MB, or by the fact that you have 3 GIGS. :eek:
  • mmonninmmonnin Centreville, VA
    edited July 2005
    For real!!
  • DoctorGeo2008DoctorGeo2008 Flint, MI
    edited July 2005
    Yeah, I only wish it was mine to take home. :tongue:
Sign In or Register to comment.