[Insight-users] Multithreading registration and Condor

Sara Rolfe smrolfe at u.washington.edu
Tue Sep 27 13:41:54 EDT 2011


Hello,

I'm attempting to run a multithreaded registration.  I am using a  
computing cluster with 13 nodes managed by Condor.  Each node us a  
dual quad-core with 32GB RAM, running the 64-bit version of RHEL.

I recompiled ITK with the CMAKE flags:
ITK_USE_REVIEW
ITK_USE_OPTIMIZED_REGISTRATION_METHODS

and added the line:
registration->SetNumberOfThreads( 8 );

to my registration code.  I've run the registration with the number of  
threads set to 1, 4 and 8.  However, I'm getting no improvement in the  
run time.  I would any advice on debugging this.  I suspect that I am  
not actually getting the additional threads.  When I look at the job  
details it looks like only one CPU was requested.

Below are the results from "condor_history -long pid" for the three  
jobs (1, 4, and 8 threads):

2076.0   1 Thread   	Run Time: 0+05:28:28
LocalUserCpu		0
LocalSysCpu		0
RemoteUserCpu		23858
RemoteSysCpu		10
RequestCpus		1

2075.0   4 threads   	Run Time: 0+05:27:41
LocalUserCpu 		0
LocalSysCpu		0
RemoteUserCpu		23680
RemoteSysCpu		9
RequestCpus		1

2077.0   8 Threads  	Run Time: 0+05:30:14
LocalUserCpu		0
LocalSysCpu		0
RemoteUserCpu		24025
RemoteSysCpu		10
RequestCpus		1

I also reran the job requesting 8 threads so I could look at the  
allocation.  The job was allocated one slot on one node. I then Iogged  
onto that node and used the "top" command and the results are below.    
Here it looks to me like multiple CPUs are being used.  I'd appreciate  
any thoughts on interpreting this.

top - 09:53:43 up 14 days, 20:45,  1 user,  load average: 2.31, 0.84,  
0.31
Tasks: 	233 total,   3 running, 230 sleeping,   0 stopped,   0 zombie
Cpu0  :  	0.0%us,  0.0%sy, 40.2%ni, 59.8%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu1  :  	0.0%us,  0.0%sy, 85.7%ni, 14.3%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu2  :  	0.0%us,  0.3%sy, 49.2%ni, 50.5%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu3  :  	0.0%us,  0.0%sy, 84.9%ni, 15.1%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu4  :  	0.0%us,  0.3%sy, 45.0%ni, 54.6%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu5  :  	0.0%us,  0.0%sy, 84.4%ni, 15.6%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu6  :  	0.0%us,  0.3%sy, 39.9%ni, 59.8%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu7  :  	0.0%us,  0.3%sy, 84.4%ni, 15.2%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu8  :  	0.0%us,  0.3%sy, 95.7%ni,  4.0%id,   0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu9  :  	0.0%us,  0.0%sy, 93.0%ni,  7.0%id,   0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu10 :  0.0%us,  0.0%sy, 85.7%ni, 14.3%id,  0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu11 :  0.0%us,  0.3%sy, 95.0%ni,  4.7%id,   0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu12 :  0.0%us,  0.0%sy, 94.7%ni,  5.3%id,   0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu13 :  0.0%us,  0.0%sy, 99.3%ni,  0.7%id,   0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu14 :  0.0%us,  0.0%sy, 92.0%ni,  8.0%id,   0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Cpu15 :  0.0%us,  0.0%sy, 96.7%ni,  3.3%id,   0.0%wa,  0.0%hi,   
0.0%si,  0.0%st
Mem:  32948672k total,  3668612k used, 29280060k free,   359324k buffers
Swap: 31262480k total,        0k used, 31262480k free,  1014788k cached

PID 		USER      PR  NI  VIRT  	RES  	SHR 	S 	%CPU %MEM    TIME+  	 
COMMAND
3333 	smrolfe    35  10   3004m 	2.0g 	3700 	R 	1267.0  	   6.4  	 
5:40.47 	condor_exec.exe
3406 	smrolfe    15    0    12880 	1216  	832   	R  	 0.3  	   0.0  	 
0:00.22 	top
1 root      	          15   0    10368  	632  	536 	S  	 0.0  	    
0.0   	0:01.79 	init
2 root      	          RT -5     0    		0    		0 		S  	 0.0  	   0.0    
	0:00.04 	migration/0
3 root     	      		  34  19    0    		0    		0 		S  	 0.0  	    
0.0        0:00.00 	ksoftirqd/0
4 root      	      	  RT  -5    0    		0    		0 		S  	 0.0  	    
0.0        0:00.00 	watchdog/0
5 root      	      	  RT  -5    0    		0    		0 		S  	 0.0  	    
0.0        0:00.03 	migration/1
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20110927/41c30472/attachment.htm>


More information about the Insight-users mailing list