Posts by Jack Shultz

1) Message boards : Number crunching : Not running feeder (Message 2683)
Posted 1 Oct 2011 by Profile Jack Shultz
Post:
apologies, I see the stat update had a period of 7 days
2) Message boards : Number crunching : Not running feeder (Message 2681)
Posted 27 Sep 2011 by Profile Jack Shultz
Post:
Hi I have just deployed a new app. I think the feeder is working but I'm debugging workunits
3) Message boards : Cafe : Nature Walks (Message 2428)
Posted 12 Jun 2010 by Profile Jack Shultz
Post:
Went to the grocery store and met DC mayor Adrian Fenty

4) Message boards : Number crunching : Linux MDrun Cuda app (Message 2315)
Posted 5 Mar 2010 by Profile Jack Shultz
Post:
Seems like every time I start trying to work on CUDA something else comes up. I'll put an effort this weekend.
5) Message boards : Number crunching : Mac Platform (Message 2307)
Posted 3 Mar 2010 by Profile Jack Shultz
Post:
Please send me a PM if you are experiencing problems with Mac. If you have a mac and some experiencing with compiling, it might be helpful if you could compile one of these applications for me.
6) Message boards : Number crunching : Help Wanted - build gpu autodock (Message 2296)
Posted 2 Mar 2010 by Profile Jack Shultz
Post:
Here is some helpful information. If you want a Visual studio solution that builds GROMACS, you can find some older versions of GROMACS here
http://www.gromacs.org/index.php?title=Download_%26_Installation/User_contributions/Gromacs_binaries/i686%2F%2FMS_Windows

Last weekend I had to update my home environment becaues my preview release of Win7 was expiring. I will try to learn more about GPU programing and see if I can modify this existing solution to use a GPU.
7) Message boards : Number crunching : UP, DOWN, CPU WU ERRORS (Message 2285)
Posted 27 Feb 2010 by Profile Jack Shultz
Post:
Sorry about the download errors. I had no internet yesterday evening. I see what the cause is now. I elimentated all the autodock workunits from the queue. To be brief, the same timestamp I use was applied to every workunit input file. This caused a problem with input files not being unique. I now have two timestamps. One applied to the batch of workunits generated that we put on the workunit name and the other generated for each ligand.
8) Message boards : Number crunching : UP, DOWN, CPU WU ERRORS (Message 2278)
Posted 24 Feb 2010 by Profile Jack Shultz
Post:
haven't found any successful 5.13 rsc... task. Not only at mine machines, even any Linux 64bit machine I have checked had the last lines


wrapper: running ./Python25/python.exe ("./MGLToolsPckgs/AutoDockTools/Utilities24/summarize_docking.py" -l out_1.dlg -r receptor_1.pdbqt -o summary_1.txt)
app exit status: 0x200
Did somebody crunch any last 5.13 autodock rcs... task by 64bit Linux successfully?

As I can remember, autodock rcs tasks (5.11 on Linux 64bit) were OK about 2 or 4 weeks ago, but I am not sure I am right.


RCS workflow runs a couple additional scripts that standard Autodock does not run. I did not catch this error with 5.13 until now. I may be able to fix this error later on today, but I'm going to check out the Terra Cotta warriors at the National Geographic museum tonight.
9) Message boards : Number crunching : UP, DOWN, CPU WU ERRORS (Message 2267)
Posted 24 Feb 2010 by Profile Jack Shultz
Post:
rcs tasks are using autodock with a modified job file. The Linux autodock platforms are currently these versions.

Linux running on an Intel x86-compatible CPU 2.25 2 Feb 2010 0:56:55 UTC
Linux running on an AMD x86_64 or Intel EM64T CPU 5.13 2 Feb 2010 1:15:44 UTC
10) Message boards : Number crunching : Help Wanted - build gpu autodock (Message 2262)
Posted 23 Feb 2010 by Profile Jack Shultz
Post:
Hello,

The cygwin environment does have 32 bit gcc compiler only. There is mingw compiler that produces 64 bit code. It may be related to your problems. I have posted the tutorial about 64bit compilation of Autodock here:
http://autodock.scripps.edu/faqs-help/how-to/how-to-compile-autodock-as-native-64-bit-windows-application


Thanks Wojciech, I will look over this the next few days. I hit a wall in my integration of OpenMM GROMACS. Andrew deffinitely wants to see a GPU app running. I'll see what we can do.

Also there is a project called GPU ROCS on SIMTK.ORG that I looked at briefly.
11) Message boards : Number crunching : UP, DOWN, CPU WU ERRORS (Message 2260)
Posted 23 Feb 2010 by Profile Jack Shultz
Post:
Send me a PM if we are still having problems with RCS. I think I cleaned it up.
12) Message boards : Number crunching : UP, DOWN, CPU WU ERRORS (Message 2258)
Posted 23 Feb 2010 by Profile Jack Shultz
Post:
The RCS workunits had some issues. I'm revising the script that generates that work. Sorry about the problem.
13) Message boards : Number crunching : Changes in WUs and tasks (Message 2235)
Posted 10 Feb 2010 by Profile Jack Shultz
Post:
I installed a vista 64 bit vmware. Its checking pointing properly after reboot. I have a suggestion that might resolve your problem. On the newer versions of the client, when you terminate the boinc manager, a box appears asking if want to exit the science applications. Please check this option. Otherwise the apps are likely to keep running unless you manually terminate those processes.
14) Message boards : Number crunching : Changes in WUs and tasks (Message 2234)
Posted 9 Feb 2010 by Profile Jack Shultz
Post:
I have tested checkpointing on xp, windows 7, linux 32, linux 64. I'm working on Snow Leopard support now. I have vista disk. When I get a chance, I'll see if I can load a VM on it.
15) Message boards : Number crunching : Project Developer Team (Message 2233)
Posted 9 Feb 2010 by Profile Jack Shultz
Post:
What's zephyr?


There are two Gromacs related projects on the SimTK site https://simtk.org/xml/index.xml

Openmm is a project that is developing GPU Api for molecular modeling
Zephyr is a project developing a gui interface that integrates the Openmm api. So you can test this Zephyr with little or no knowledge of molecular dynamics.

I think this will be much easier if I find a way to get the DLLs working in the current working directory of the application.
16) Message boards : Number crunching : Changes in WUs and tasks (Message 2232)
Posted 9 Feb 2010 by Profile Jack Shultz
Post:
Please check that the job file has these options

<application>mdrun.exe</application>
<command_line> -v -x -c md.pdb -o -e -cpi md.cpt -cpt 1 -deffnm md</comm
and_line>

These are the options that have been successful for me.
So how do you check on the job file options? First find out which slot directory your workunit occupies. Select the workunit and click properties.

mine says slot\4

On my version of BOINC and Windows 7 I can find slot 4 under
C:\ProgramData\BOINC\slots\4

Open job.xml and you will see which command_line options are used for mdrun.exe
17) Message boards : Number crunching : Checkpoint on reboot (Message 2227)
Posted 8 Feb 2010 by Profile Jack Shultz
Post:
checkpointing for autodock is another issue. Less important because users will loose much less credit.
18) Message boards : Number crunching : Checkpoint on reboot (Message 2226)
Posted 8 Feb 2010 by Profile Jack Shultz
Post:
Basically, I'm trying not to reinvent the wheel. The application has its own checkpointing options. You can read about those checkpoint options here.
http://www.psc.edu/general/software/packages/gromacs/online/mdrun.html

I know as a mater of fact we can continue mdrun from where we left off via command line. Running it with the wrapper is not working the way we intend. I spent some time trying to fix a problem with the fraction complete reporting. I fixed that but its still not restarting where we left off when I restart boinc.
19) Message boards : Number crunching : Changes in WUs and tasks (Message 2225)
Posted 8 Feb 2010 by Profile Jack Shultz
Post:
I have tested the changes on windows and it seems to checkpoint properly. Testing it on linux now. Its not behaving the way I expected. It does not use the state.cpt file. Instead it starts the simulation from the beginning. When I run the same command from the terminal it continues where it left off. So something is not making sense here.
20) Message boards : Cafe : Nature Walks (Message 2224)
Posted 8 Feb 2010 by Profile Jack Shultz
Post:
lost power at home yesterday. glad I don't host @home


Next 20


©2017 All rights reserved | Design by Digital BioPharm Ltd