Windows MDrun Cuda app

Message boards : Number crunching : Windows MDrun Cuda app

To post messages, you must log in.

1 · 2 · 3 · Next

AuthorMessage
Profile nenym

Send message
Joined: 23 Apr 09
Posts: 99
Credit: 530,784
RAC: 985
Message 1196 - Posted: 8 Sep 2009, 4:30:07 UTC
Last modified: 8 Sep 2009, 4:34:55 UTC

Strange tasks.

100 % done at 1s, then changing task, the same over queue, then returned to the first task and running and running when 100% done....no GPU load, 13% CPU load.
Delete or is there any chance to finish these tasks?

mdurn with cuda 0.26 md_P10_TYR52...
The same at hosts 169 (XP x64 + GTX260) and 211 (XP x86 + 9600GT).
ID: 1196 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Rabinovitch
Avatar

Send message
Joined: 22 Apr 09
Posts: 47
Credit: 82,303
RAC: 0
Message 1197 - Posted: 8 Sep 2009, 5:29:21 UTC
Last modified: 8 Sep 2009, 5:39:50 UTC

In my case 100% of task was indeed achieved in 1 s, but already 2 hours this task is still running and seems to be OK - only time will show the result.

Unfortunately, this WU uses 1 full CPU core, just like in Einstein@h. I hope that it's only temporary bug...

Win 7 x64. BM 6.10.3

p.s. mdrun_openmm.exe crashed, as a result:

08.09.2009 12:35:18 DrugDiscovery [error] Can't rename output file md_P10_TYR52_LEU33_GLN56_ga10_34929_ChemDiv_8017-0401_125183574246522681_1252088741829242196_3_0 to projects/boinc.drugdiscoveryathome.com/md_P10_TYR52_LEU33_GLN56_ga10_34929_ChemDiv_8017-0401_125183574246522681_1252088741
08.09.2009 12:35:18 DrugDiscovery Computation for task md_P10_TYR52_LEU33_GLN56_ga10_34929_ChemDiv_8017-0401_125183574246522681_1252088741829242196_3 finished

It was being processed during 2 h 30 m.
ID: 1197 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile nenym

Send message
Joined: 23 Apr 09
Posts: 99
Credit: 530,784
RAC: 985
Message 1198 - Posted: 8 Sep 2009, 6:36:36 UTC

Hmm....aborting all CUDA 0.26 tasks.
ID: 1198 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Inais
Avatar

Send message
Joined: 1 May 09
Posts: 2
Credit: 23,685
RAC: 0
Message 1199 - Posted: 8 Sep 2009, 9:03:05 UTC

crashed with Error Code 108

Name md_P10_TYR52_LEU33_GLN56_ga10_34052_ChemDiv_8012-5936__1252373357727423743_2
Workunit 113246
Created 8 Sep 2009 1:29:19 UTC
Sent 8 Sep 2009 8:55:35 UTC
Received 8 Sep 2009 8:57:07 UTC
Server state Over
Outcome Client error
Client state Compute error
Exit status -108 (0xffffff94)
I wish I can fly like a bird in the sky .....
ID: 1199 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile nenym

Send message
Joined: 23 Apr 09
Posts: 99
Credit: 530,784
RAC: 985
Message 1200 - Posted: 8 Sep 2009, 11:01:30 UTC

Task as a toddler.
It needs 0.25 of one CPU core, full CUDA and one 24/7 volunteer's baby-sitting (necessary to suspend all other CUDA tasks include DD CUDA tasks) to crash after long time....
ID: 1200 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Rabinovitch
Avatar

Send message
Joined: 22 Apr 09
Posts: 47
Credit: 82,303
RAC: 0
Message 1201 - Posted: 8 Sep 2009, 12:02:09 UTC

Second cuda 0.25 WU crashed because of mdrun_openmm.exe fault. And again after 02:53:29! Seems like it's common bug, and I abort all this tasks...
ID: 1201 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile nenym

Send message
Joined: 23 Apr 09
Posts: 99
Credit: 530,784
RAC: 985
Message 1202 - Posted: 8 Sep 2009, 13:57:09 UTC - in response to Message 1201.  
Last modified: 8 Sep 2009, 14:15:48 UTC

The same at me, after 2:55:52 crashed.

08/09/2009 15:45:02	DrugDiscovery	[error] Can't rename output file md_P10_TYR52_LEU33_GLN56_ga10_34052_ChemDiv_8012-5936__1252373448848333509_0_0 to projects/boinc.drugdiscoveryathome.com/md_P10_TYR52_LEU33_GLN56_ga10_34052_ChemDiv_8012-5936__1252373448848333509_0_0: Error 2
08/09/2009 15:45:02	DrugDiscovery	Computation for task md_P10_TYR52_LEU33_GLN56_ga10_34052_ChemDiv_8012-5936__1252373448848333509_0 finished

Maybe I am a bad baby-sitter. The child died.
<![CDATA[
<message>
 - exit code 195 (0xc3)
</message>
<stderr_txt>
wrapper: starting
12:49:05 (3964): wrapper: running mdrun_openmm.exe (-v -deffnm md -v -deffnm md --device 0)
app exit status: 0xc0000005
15:44:54 (3964): called boinc_finish
</stderr_txt>
]]>
ID: 1202 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Tim Turner
Avatar

Send message
Joined: 1 May 09
Posts: 570
Credit: 184,322
RAC: 0
Message 1203 - Posted: 8 Sep 2009, 14:05:10 UTC - in response to Message 1202.  
Last modified: 8 Sep 2009, 14:56:15 UTC

well, it's good to see jack has gotten the windows version off the ground, although it seems that the windows and linux needs some work.

i'm running one right now, and i see that checkpoints aren't in this version of the app.
Tim Turner
Public Relations Admin
Secunia PSI: http://secunia.com/vulnerability_scanning/personal/
If you need help via voice or Convo; PM me and i will give you details on where i will be; Teamspeak, Yahoo Messenger, or Skype.
ID: 1203 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile [AF>EDLS] frederic abussan
Avatar

Send message
Joined: 1 May 09
Posts: 30
Credit: 849,263
RAC: 0
Message 1205 - Posted: 8 Sep 2009, 15:32:34 UTC - in response to Message 1203.  

Yes
i got mi first Cuda wu windows on ID 353, lets se wat she can do
Here today gone tomorrow
ID: 1205 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Tim Turner
Avatar

Send message
Joined: 1 May 09
Posts: 570
Credit: 184,322
RAC: 0
Message 1206 - Posted: 8 Sep 2009, 15:41:49 UTC - in response to Message 1205.  

right now, the error rate of the cuda version for windows is 100 %. that being said, jack is going to try to figure it out later tonight after he gets off of work.

It seems that their is trouble further down the line in the wu's towards the
end-of-life of the wu.
my wu has gone 1:34 hr/minutes. it'll error out at about 2 i think based on earlier post.
Tim Turner
Public Relations Admin
Secunia PSI: http://secunia.com/vulnerability_scanning/personal/
If you need help via voice or Convo; PM me and i will give you details on where i will be; Teamspeak, Yahoo Messenger, or Skype.
ID: 1206 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile nenym

Send message
Joined: 23 Apr 09
Posts: 99
Credit: 530,784
RAC: 985
Message 1208 - Posted: 8 Sep 2009, 16:22:37 UTC - in response to Message 1205.  
Last modified: 8 Sep 2009, 16:22:57 UTC

"frederic" wrote:
Yes
i got mi first Cuda wu windows on ID 353, lets se wat she can do

No chance to finish DD CUDA task, switch to GPUGRID (Collatz's server is down, Seti and AQUA out of work...).
ID: 1208 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Tim Turner
Avatar

Send message
Joined: 1 May 09
Posts: 570
Credit: 184,322
RAC: 0
Message 1209 - Posted: 8 Sep 2009, 16:27:21 UTC - in response to Message 1208.  

yea, i'm still on my first one, made some recommendations to jack about the app.

it seems that in the past week, every gpu projects has had trouble in some form or fashion.
Tim Turner
Public Relations Admin
Secunia PSI: http://secunia.com/vulnerability_scanning/personal/
If you need help via voice or Convo; PM me and i will give you details on where i will be; Teamspeak, Yahoo Messenger, or Skype.
ID: 1209 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile [AF>EDLS] frederic abussan
Avatar

Send message
Joined: 1 May 09
Posts: 30
Credit: 849,263
RAC: 0
Message 1211 - Posted: 8 Sep 2009, 16:44:47 UTC - in response to Message 1209.  

Merci beaucoup

Tank's a lot for the explain
Here today gone tomorrow
ID: 1211 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Morgan the Gold

Send message
Joined: 27 Aug 09
Posts: 4
Credit: 59,854
RAC: 112
Message 1212 - Posted: 8 Sep 2009, 18:22:00 UTC

:D got cuda tasks,
:( [error] can't rename output file md_...
trying reset project..
:( no go

server 2003 64bit beta, ingenuous.
ID: 1212 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Tim Turner
Avatar

Send message
Joined: 1 May 09
Posts: 570
Credit: 184,322
RAC: 0
Message 1213 - Posted: 8 Sep 2009, 18:59:38 UTC - in response to Message 1212.  

:D got cuda tasks,
:( [error] can't rename output file md_...
trying reset project..
:( no go

server 2003 64bit beta, ingenuous.


don't bother, jack's going to try to figure it out later tonight if possible. and he'll try to work on a couple other things that are need to be a successful cuda app.

Tim Turner
Public Relations Admin
Secunia PSI: http://secunia.com/vulnerability_scanning/personal/
If you need help via voice or Convo; PM me and i will give you details on where i will be; Teamspeak, Yahoo Messenger, or Skype.
ID: 1213 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile nenym

Send message
Joined: 23 Apr 09
Posts: 99
Credit: 530,784
RAC: 985
Message 1215 - Posted: 8 Sep 2009, 21:50:06 UTC
Last modified: 8 Sep 2009, 21:58:44 UTC

Current CUDA task (21:34:57 UTC) could be OK or is it the old one?
ID: 1215 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Tim Turner
Avatar

Send message
Joined: 1 May 09
Posts: 570
Credit: 184,322
RAC: 0
Message 1216 - Posted: 8 Sep 2009, 22:00:48 UTC - in response to Message 1215.  

jack said that he released a batch this afternoon with a couple of parameters set differently, so hope this works.
Tim Turner
Public Relations Admin
Secunia PSI: http://secunia.com/vulnerability_scanning/personal/
If you need help via voice or Convo; PM me and i will give you details on where i will be; Teamspeak, Yahoo Messenger, or Skype.
ID: 1216 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile nenym

Send message
Joined: 23 Apr 09
Posts: 99
Credit: 530,784
RAC: 985
Message 1217 - Posted: 8 Sep 2009, 22:08:57 UTC - in response to Message 1216.  

OK, I try.
ID: 1217 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Tim Turner
Avatar

Send message
Joined: 1 May 09
Posts: 570
Credit: 184,322
RAC: 0
Message 1218 - Posted: 8 Sep 2009, 22:44:22 UTC - in response to Message 1217.  

he may cancel the old and new ones, it's a tough process to do in the control panel.
Tim Turner
Public Relations Admin
Secunia PSI: http://secunia.com/vulnerability_scanning/personal/
If you need help via voice or Convo; PM me and i will give you details on where i will be; Teamspeak, Yahoo Messenger, or Skype.
ID: 1218 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile nenym

Send message
Joined: 23 Apr 09
Posts: 99
Credit: 530,784
RAC: 985
Message 1220 - Posted: 9 Sep 2009, 13:46:05 UTC
Last modified: 9 Sep 2009, 13:51:03 UTC

mdur with cuda 0.30 (cuda23) 03:06:14 elapsed time and still running (XP x86, 9600GTX host ID 211).
What time can I estimate? Before starting that task was estimated time 38hours, is it right? If it is, deadline 12 sept. 2009 is too soon.
ID: 1220 · Rating: 0 · rate: Rate + / Rate - Report as offensive
1 · 2 · 3 · Next

Message boards : Number crunching : Windows MDrun Cuda app


©2017 All rights reserved | Design by Digital BioPharm Ltd