[GE users] [sge6] Jobs submitted succesfully won't run.

Stephan Grell stephan.grell at sun.com
Sat Jul 24 14:30:59 BST 2004


Wee Yeh Tan wrote:

>On Sat, 24 Jul 2004 12:58:17 +0200, Reuti <reuti at staff.uni-marburg.de> wrote:
>  
>
>>>>qhost
>>>>        
>>>>
>>Anything interesting in:
>>
>>qhost -F or qstat -f
>>    
>>
>
>Nothing here...  This is a sample output in case I missed something...
>all.q at comp3.cl  BIP   0/2       0.00     lx24-amd64    
>        hl:load_avg=0.000000
>        hl:load_short=0.000000
>        hl:load_medium=0.000000
>        hl:load_long=0.000000
>        hl:arch=lx24-amd64
>        hl:num_proc=2.000000
>        hl:mem_free=1.842G
>        hl:swap_free=1.999G
>        hl:virtual_free=3.841G
>        hl:mem_total=1.967G
>        hl:swap_total=1.999G
>        hl:virtual_total=3.966G
>        hl:mem_used=128.062M
>        hl:swap_used=0.000
>        hl:virtual_used=128.062M
>        hl:cpu=0.000000
>        hl:np_load_avg=0.000000
>        hl:np_load_short=0.000000
>        hl:np_load_medium=0.000000
>        hl:np_load_long=0.000000
>        qf:qname=all.q
>        qf:hostname=comp3.cl
>        qc:slots=2.000000
>        qf:tmpdir=/tmp
>        qf:seq_no=0.000000
>        qf:rerun=0.000000
>        qf:calendar=NONE
>        qf:s_rt=infinity
>        qf:h_rt=infinity
>        qf:s_cpu=infinity
>        qf:h_cpu=infinity
>        qf:s_fsize=infinity
>        qf:h_fsize=infinity
>        qf:s_data=infinity
>        qf:h_data=infinity
>        qf:s_stack=infinity
>        qf:h_stack=infinity
>        qf:s_core=infinity
>        qf:h_core=infinity
>        qf:s_rss=infinity
>        qf:h_rss=infinity
>        qf:s_vmem=infinity
>        qf:h_vmem=infinity
>        qf:min_cpu_interval=00:05:00
>
>  
>
>>to see all of the values? Is a simple qrsh working to get an interactive login?
>>    
>>
>
>The qrsh job got stuck like the rest of 'em..
>
>I'm just wondering if there is anyway I can pry deeper into the system
>to see what SGE is doing.  Is there anyway to turn up the verbosity of
>qmaster?
>  
>
Well, yes, there is. But you are interested in the scheduler, since it 
is not dispatching your jobs.

There are several different ways. One is to turn on the profiling 
information, with "params profile=1"
in the scheduler configurations (qconf -msconf). The output is loged in 
the scheduler messages file. The
problem is, that there is no description for the profiling output yet, 
and we do change the format every once
in a while.

Well, you can also run the scheduler in debug mode. You can source the 
$SGE_ROOT/util/dl1.csh script.
If you parse 0 into it, the debug mode is turned off,
1 gives you some output
2 gives you pretty much everything.

Can you post the scheduler configuraton?

Had you have a look into the scheduler message file? 
$SGE_ROOT/<cell>/spool/qmaster/schedd/messages

I hope it helps.
Stephan


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list