[Wien] Query regarding running less cpu core than available in the machine

[Gavin Abo]

FYI,

If you have "top" on your system and want to instead see the CPU usage 
of each core, refer to [1].

1. Run the command: top
2. Then, try pressing: 1

On the other hand, it seems not easy to monitor the thread usage of each 
core since the thread might switch cores for thermal load balancing [2].

It looks to me too that you might have provided the CPU overall capacity 
of each lapw1 process. The Irix mode might have to be toggled in top 
using I (from OFF for CPU overall capacity to ON for CPU thread) [3].

With top running, it looks like the total thread count can be obtained 
using [4]: H

If you want to see the number threads being using by each process (e.g., 
lapw1), nTH might be selected for that.  Refer to [5,6].

1. With top running, press: f
2. Use arrow keys to select nTH, then press: d
3. To quit field management window, press: q

If your using a cluster having a monitoring tool, like Ganglia [7], I 
suggest consulting the documentation for your monitoring tool to see if 
it displays the CPU and threading information.

If your using a cluster with a qsub queue system, there are programs 
such as qstat and pdsh [8].  Though, if you are using qsub or a 
different queue system like bsub, consult the documentation for your 
particular system (e.g., man pages [9]).

[1] https://bencane.com/2011/08/28/top-show-cpu-usage-on-all-cores/
[2] 
https://bitsum.com/tips-and-tweaks/what-is-the-difference-between-a-thread-and-a-core/
[3] 
https://stackoverflow.com/questions/3342889/how-do-i-measure-separate-cpu-core-usage-for-a-process
[4] http://ask.xmodulo.com/view-threads-process-linux.html
[5] 
https://superuser.com/questions/1123705/using-linux-top-how-do-i-view-the-thread-count-of-a-process
[6] 
https://stackoverflow.com/questions/8032372/how-can-i-see-which-cpu-core-a-thread-is-running-in
[7] https://en.wikipedia.org/wiki/Ganglia_(software)
[8] 
https://www.osc.edu/supercomputing/batch-processing-at-osc/monitoring-and-managing-your-job
[9] https://en.wikipedia.org/wiki/Man_page


On 7/8/2019 1:03 PM, Peter Blaha wrote:
> You have to explain exactly what you are doing and what the jobs do or 
> not.
>
> a) How did you compile wien2k ? ifort or gfortran ??? Which blas ?
>
> b) I do not understand your statement:
>
>>                 I have seen it directly from cpu monitor. The usage 
>> of the cpus are 100% at all the 48 cpus (24 core). I have started 4 
>> calculations & the cpu usage are as follows:
>> lapw1  2400
>> lapw1  600
>> lapw1  400
>> lapw1  100
>
> So you had 24 cores, 6 cores, 4 cores and only 1 core on the 4 jobs ???
> They are all different, aren't they ??
>
> Why are you saying they use 48 (or 24 cores) ???
>
> What is "cpu monitor" ???
>
> What are the case.dayfile outputs telling you ???
>
> I'd like to see the timing lines in the dayfiles.
>
> The timing should be different for these 4 jobs ....
>
>
>>
>> Eagerly waiting for your advice.
>>
>> with regards,
>>
>> On Mon, 8 Jul 2019 at 21:33, shamik chakrabarti 
>> <shamik15041981 at gmail.com <mailto:shamik15041981 at gmail.com>> wrote:
>>
>>     Dear Sir,
>>
>>                     I have seen it directly from cpu monitor. The usage
>>     of the cpus are 100% at all the 48 cpus (24 core)
>>
>>     On Mon, 8 Jul 2019 at 21:11, Peter Blaha
>>     <pblaha at theochem.tuwien.ac.at <mailto:pblaha at theochem.tuwien.ac.at>>
>>     wrote:
>>
>>         How did you find out that it has taken all 24 cores ?
>>
>>         Am 08.07.2019 um 13:06 schrieb shamik chakrabarti:
>>          > Dear wien2k users,
>>          >
>>          >                               I have installed wien2k in a
>>         system having
>>          > Ubuntu. If I set OMP_NUM_THREAD = 2/3/4, it is taking all the
>>         available
>>          > 24 cores. Whereas if I set OMP_NUM_THREAD = 1 it is taking
>>         only 1 cpu
>>          > per calculation. My query is how can I use more than 1 cpu
>>         but less than
>>          > the total available cores (say I want to use 20 cores while
>>         total
>>          > available cores are 24)..
>>          >
>>          > Any response in this regard will be helpful for us.
>>          >
>>          > with regards,
>>          >
>>          > --
>>          > Dr. Shamik Chakrabarti
>>          > Project Employee
>>          > Dept. of Metallurgical and Materials Engineering
>>          > IIT Kharagpur
>>          > Kharagpur-721302
>>          > INDIA
>>          >
>>          > _______________________________________________
>>          > Wien mailing list
>>          > Wien at zeus.theochem.tuwien.ac.at
>>         <mailto:Wien at zeus.theochem.tuwien.ac.at>
>>          > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>          > SEARCH the MAILING-LIST at:
>> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>>          >
>>
>>         -- 
>> --------------------------------------------------------------------------
>>         Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 
>> Vienna
>>         Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
>>         Email: blaha at theochem.tuwien.ac.at
>>         <mailto:blaha at theochem.tuwien.ac.at>    WIEN2k: 
>> http://www.wien2k.at
>>         WWW:
>> http://www.imc.tuwien.ac.at/tc_blaha-------------------------------------------------------------------------
>>
>>
>>         _______________________________________________
>>         Wien mailing list
>>         Wien at zeus.theochem.tuwien.ac.at
>>         <mailto:Wien at zeus.theochem.tuwien.ac.at>
>>         http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>         SEARCH the MAILING-LIST at:
>> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>>
>>
>>
>>     --     Dr. Shamik Chakrabarti
>>     Project Employee
>>     Dept. of Metallurgical and Materials Engineering
>>     IIT Kharagpur
>>     Kharagpur-721302
>>     INDIA
>>
>>
>>
>> -- 
>> Dr. Shamik Chakrabarti
>> Project Employee
>> Dept. of Metallurgical and Materials Engineering
>> IIT Kharagpur
>> Kharagpur-721302
>> INDIA
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20190708/818b020c/attachment.html>