Page 1 of 1
Project: 6903 (Run 10, Clone 5, Gen 114)
Posted: Wed May 09, 2012 5:43 pm
by amang
Not sure what happened with this WU, but it keeps failing on my rig.
Code: Select all
# Linux SMP Console Edition ###################################################
###############################################################################
Folding@Home Client Version 6.34
http://folding.stanford.edu
###############################################################################
###############################################################################
Launch directory: /usr/local/fah
Executable: ./fah6
Arguments: -bigadv -smp
[17:22:43] - Ask before connecting: No
[17:22:43] - User name: Amang (Team 37726)
[17:22:43] - User ID: E414B2146643772
[17:22:43] - Machine ID: 1
[17:22:43]
[17:22:43] Work directory not found. Creating...
[17:22:43] Could not open work queue, generating new queue...
[17:22:43] - Preparing to get new work unit...
[17:22:43] Cleaning up work directory
[17:22:43] + Attempting to get work packet
[17:22:43] Passkey found
[17:22:43] - Connecting to assignment server
[17:22:44] - Successful: assigned to (130.237.232.237).
[17:22:44] + News From Folding@Home: Welcome to Folding@Home
[17:22:44] Loaded queue successfully.
[17:30:05] + Closed connections
[17:30:05]
[17:30:05] + Processing work unit
[17:30:05] Core required: FahCore_a5.exe
[17:30:05] Core found.
[17:30:05] Working on queue slot 03 [May 9 17:30:05 UTC]
[17:30:05] + Working ...
[17:30:05]
[17:30:05] *------------------------------*
[17:30:05] Folding@Home Gromacs SMP Core
[17:30:05] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[17:30:05]
[17:30:05] Preparing to commence simulation
[17:30:05] - Looking at optimizations...
[17:30:05] - Created dyn
[17:30:05] - Files status OK
[17:30:08] - Expanded 57247356 -> 71846524 (decompressed 50.4 percent)
[17:30:08] Called DecompressByteArray: compressed_data_size=57247356 data_size=71846524, decompressed_data_size=71846524 diff=0
[17:30:09] - Digital signature verified
[17:30:09]
[17:30:09] Project: 6903 (Run 10, Clone 5, Gen 114)
[17:30:09]
[17:30:09] Assembly optimizations on if available.
[17:30:09] Entering M.D.
[17:30:16] Mapping NT from 12 to 12
[17:30:28] Completed 0 out of 250000 steps (0%)
[17:31:19] + Closed connections
[17:31:19]
[17:31:19] + Processing work unit
[17:31:19] Core required: FahCore_a5.exe
[17:31:19] Core found.
[17:31:19] Working on queue slot 01 [May 9 17:31:19 UTC]
[17:31:19] + Working ...
[17:31:20]
[17:31:20] *------------------------------*
[17:31:20] Folding@Home Gromacs SMP Core
[17:31:20] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[17:31:20]
[17:31:20] Preparing to commence simulation
[17:31:20] - Ensuring status. Please wait.
[17:31:29] - Looking at optimizations...
[17:31:29] - Working with standard loops on this execution.
[17:31:29] - Created dyn
[17:31:29] - Files status OK
[17:31:38] - Expanded 57247356 -> 71846524 (decompressed 50.4 percent)
[17:31:38] Called DecompressByteArray: compressed_data_size=57247356 data_size=71846524, decompressed_data_size=71846524 diff=0
[17:31:38] - Digital signature verified
[17:31:38]
[17:31:38] Project: 6903 (Run 10, Clone 5, Gen 114)
[17:31:38]
[17:31:38] Entering M.D.
[17:31:46] Mapping NT from 12 to 12
[17:31:54] CoreStatus = 0 (0)
[17:31:54] CoreStatus = 0 (0)
[17:31:54] Sending work to server
[17:31:54] Project: 6903 (Run 10, Clone 5, Gen 114)
[17:31:54] - Error: Could not get length of results file work/wuresults_03.dat
[17:31:54] - Error: Could not read unit 03 file. Removing from queue.
[17:31:54] - Preparing to get new work unit...
[17:31:54] Cleaning up work directory
[17:31:54] + Attempting to get work packet
[17:31:54] Passkey found
[17:31:54] Sending work to server
[17:31:54] Project: 6903 (Run 10, Clone 5, Gen 114)
[17:31:54] - Error: Could not get length of results file work/wuresults_01.dat
[17:31:54] - Error: Could not read unit 01 file. Removing from queue.
[17:31:54] - Preparing to get new work unit...
[17:31:54] Cleaning up work directory
[17:31:54] + Attempting to get work packet
[17:31:54] Passkey found
[17:31:54] - Connecting to assignment server
[17:31:54] - Connecting to assignment server
[17:31:55] - Successful: assigned to (130.237.232.237).
[17:31:55] + News From Folding@Home: Welcome to Folding@Home
[17:31:55] - Successful: assigned to (130.237.232.237).
[17:31:55] + News From Folding@Home: Welcome to Folding@Home
[17:31:55] Loaded queue successfully.
My console screen spits out some error messages like "Out of Memory" and "Core Status = 0 (0)"
I am currently running this folder on a Linux via Virtual Box. I have reserved 3GB worth of memory for this process, so it's definitely not a problem with memory.
Any input is much appreciated here.
Re: Project: 6903 (Run 10, Clone 5, Gen 114)
Posted: Wed May 09, 2012 7:02 pm
by bruce
Welcome to foldingforum.org, amang.
The out-of-memory message is probably correct. What makes you think that 3gb is enough?
The recommendations for SMP is .5 to 1 GB per core so the MINIMUM requirement for a 12-core machine would be 6-GB inside of the VM, and of course it probably takes a bit more than that to actually run the VM. You're asking for BIGadv assignments and they really are big.
We continually get requests for FAH on clusters, and although that doesn't seem to be feasible on the hardware that folks have proposed, the -bigadv assignments are specifically designed for hardware that was called a supercomputer not too long ago. For the rest of us, -smp is properly sized for anything between about 4 and 16 cores.
Re: Project: 6903 (Run 10, Clone 5, Gen 114)
Posted: Thu May 10, 2012 1:34 am
by amang
Hello Bruce. Your point is taken. I will give it another try with bigger memory size and report my finding here. Thanks!
Re: Project: 6903 (Run 10, Clone 5, Gen 114)
Posted: Thu May 10, 2012 4:33 am
by amang
After upping the memory size, the folding now works normally. Current TPF is around 1:27.
Here is the latest log:
Code: Select all
# Linux SMP Console Edition ###################################################
###############################################################################
Folding@Home Client Version 6.34
http://folding.stanford.edu
###############################################################################
###############################################################################
Launch directory: /usr/local/fah
Executable: ./fah6
Arguments: -bigadv -smp
[01:38:00] - Ask before connecting: No
[01:38:00] - User name: Amang (Team 37726)
[01:38:00] - User ID: E414B2146643772
[01:38:00] - Machine ID: 1
[01:38:00]
[01:38:00] Work directory not found. Creating...
[01:38:01] Loaded queue successfully.
[01:38:01] - Preparing to get new work unit...
[01:38:01] Cleaning up work directory
[01:38:01] + Attempting to get work packet
[01:38:01] Passkey found
[01:38:01] - Connecting to assignment server
[01:38:01] - Successful: assigned to (130.237.232.237).
[01:38:01] + News From Folding@Home: Welcome to Folding@Home
[01:38:02] Loaded queue successfully.
[01:50:57] + Closed connections
[01:50:57]
[01:50:57] + Processing work unit
[01:50:57] Core required: FahCore_a5.exe
[01:50:57] Core found.
[01:50:57] Working on queue slot 05 [May 10 01:50:57 UTC]
[01:50:57] + Working ...
[01:50:57]
[01:50:57] *------------------------------*
[01:50:57] Folding@Home Gromacs SMP Core
[01:50:57] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[01:50:57]
[01:50:57] Preparing to commence simulation
[01:50:57] - Looking at optimizations...
[01:50:57] - Created dyn
[01:50:57] - Files status OK
[01:51:00] - Expanded 57247356 -> 71846524 (decompressed 50.4 percent)
[01:51:00] Called DecompressByteArray: compressed_data_size=57247356 data_size=71846524, decompressed_data_size=71846524 diff=0
[01:51:01] - Digital signature verified
[01:51:01]
[01:51:01] Project: 6903 (Run 10, Clone 5, Gen 114)
[01:51:01]
[01:51:01] Assembly optimizations on if available.
[01:51:01] Entering M.D.
[01:51:08] Mapping NT from 12 to 12
[01:51:17] Completed 0 out of 250000 steps (0%)
[01:53:42] + Closed connections
[01:53:47]
[01:53:47] + Processing work unit
[01:53:47] Core required: FahCore_a5.exe
[01:53:47] Core found.
[01:53:47] Working on queue slot 05 [May 10 01:53:47 UTC]
[01:53:47] + Working ...
[01:53:47]
[01:53:47] *------------------------------*
[01:53:47] Folding@Home Gromacs SMP Core
[01:53:47] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[01:53:47]
[01:53:47] Preparing to commence simulation
[01:53:47] - Ensuring status. Please wait.
[01:53:57] - Looking at optimizations...
[01:53:57] - Working with standard loops on this execution.
[01:53:57] - Previous termination of core was improper.
[01:53:57] - Files status OK
[01:54:03] - Expanded 57247356 -> 71846524 (decompressed 50.4 percent)
[01:54:03] Called DecompressByteArray: compressed_data_size=57247356 data_size=71846524, decompressed_data_size=71846524 diff=0
[01:54:03] - Digital signature verified
[01:54:03]
[01:54:03] Project: 6903 (Run 10, Clone 5, Gen 114)
[01:54:03]
[01:54:04] Entering M.D.
[01:54:13] Mapping NT from 12 to 12
[01:54:37] Completed 0 out of 250000 steps (0%)
[03:21:52] Completed 2500 out of 250000 steps (1%)
[03:28:06] Completed 2500 out of 250000 steps (1%)
HFM is currently reporting a PPD of 3747.50 with 22,706 in Credit. Does this look normal?
Re: Project: 6903 (Run 10, Clone 5, Gen 114)
Posted: Thu May 10, 2012 5:59 am
by ChelseaOilman
It doesn't look normal for a computer qualified to do bigadv. Frame times on my 4P computer doing a 6903 WU are less than 11:30. An hour and a half TPF looks more like it's being folded on an overclocked 2600K system. Your going to exceed the preferred deadline which means you won't get any bonus points and the WU will get reassigned to someone else before you finish it, which is a waste of resources. If you can't beat the preferred deadline your better off doing regular SMP. At least you'll get bonus points.
Re: Project: 6903 (Run 10, Clone 5, Gen 114)
Posted: Thu May 10, 2012 6:19 am
by amang
So should I dump this WU for now? Problem is that when I delete my existing 'work folder, 'queue.dat', and 'unitinfo.txt', and redownload a new WU, this very same WU keeps coming back to me.
Or should I temporarily switch from -bigadv to -advmethods?
Re: Project: 6903 (Run 10, Clone 5, Gen 114)
Posted: Thu May 10, 2012 8:32 am
by bollix47
amang wrote:So should I dump this WU for now? Problem is that when I delete my existing 'work folder, 'queue.dat', and 'unitinfo.txt', and redownload a new WU, this very same WU keeps coming back to me.
Or should I temporarily switch from -bigadv to -advmethods?
In addition to the files you've deleted you also need to delete machinedependent.dat.
Just remove -bigadv from your options.