subject:"\[OMPI users\] Help with Specific Binding"

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Ralph Castain

Checkout the man page “OMPI_Affinity_str” for an MPI extension that might help


> On Sep 13, 2015, at 7:28 AM, Saliya Ekanayake  wrote:
> 
> Thank you, I'll try this. Also, is there a way to know which core a process 
> is bound to within the program other than executing something like taskset 
>  from program?
> 
> On Sun, Sep 13, 2015 at 10:05 AM, Ralph Castain  > wrote:
> Actually, the error was correct - it was me that was incorrect. The correct 
> set of options would be:
> 
> —map-by ppr:12_node —bind-to core —cpu-set=0,2,4,…
> 
> Sorry about the confusion
> 
> 
>> On Sep 13, 2015, at 2:43 AM, Ralph Castain > > wrote:
>> 
>> The rankfile will certainly do it, but that error is a bug and I’ll have to 
>> fix it.
>> 
>> 
>>> On Sep 13, 2015, at 1:10 AM, Saliya Ekanayake >> > wrote:
>>> 
>>> I could get it working by manually generating a rankfile all the ranks and 
>>> not using any --map-by options.
>>> 
>>> I'll try the --map-by core as well
>>> 
>>> On Sun, Sep 13, 2015 at 3:59 AM, Tobias Kloeffel >> > wrote:
>>> Hi,
>>> use: --map-by core
>>> 
>>> regards,
>>> Tobias
>>> 
>>> 
>>> On 09/13/2015 09:41 AM, Saliya Ekanayake wrote:
 I tried,
 
  --map-by ppr:12:node --slot-list 0,2,4,6,8,10,12,14,16,18,20,22 --bind-to 
 core -np 12
 
 but it complains,
 
 "Conflicting directives for binding policy are causing the policy
 to be redefined:
 
   New policy:   socket
   Prior policy:  CORE
 
 Please check that only one policy is defined.
 "
 
 On Sun, Sep 13, 2015 at 2:57 AM, Ralph Castain > wrote:
 Try something like this instead:
 
 —map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…
 
 You’ll have to play a bit with the core numbers in the slot-list to get 
 the numbering right as I don’t know how your machine numbers them, and I 
 can’t guarantee it will work - but it’s worth a shot. If it doesn’t, then 
 I may have to add an option for such purposes
 
 Ralph
 
> On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake  > wrote:
> 
> Hi,
> 
> We've a machine as in the following picture. I'd like to run 12 MPI procs 
> per node each bound to 1 core, but like shown in blue dots in the 
> pictures. I can use the following command to run 12 procs per node, but 
> PE=1 makes all the 12 processes will run in just 1 socket. PE=2 will make 
> a process bind to 2 cores, which is not what I want. 
> 
> --map-by ppr:12:node:PE=1,SPAN
> 
> Thank you,
> Saliya
> 
> 
> 
> -- 
> Saliya Ekanayake
> Ph.D. Candidate | Research Assistant
> School of Informatics and Computing | Digital Science Center
> Indiana University, Bloomington
> Cell 812-391-4914 
>  http://saliya.org 
> ___
> users mailing list
> us...@open-mpi.org 
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
> 
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/09/27558.php 
> 
 
 ___
 users mailing list
 us...@open-mpi.org 
 Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
 
 Link to this post: 
 http://www.open-mpi.org/community/lists/users/2015/09/27559.php 
 
 
 
 
 -- 
 Saliya Ekanayake
 Ph.D. Candidate | Research Assistant
 School of Informatics and Computing | Digital Science Center
 Indiana University, Bloomington
 Cell 812-391-4914 
 http://saliya.org 
 
 ___
 users mailing list
 us...@open-mpi.org 
 Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
 
 Link to this post: 
 http://www.open-mpi.org/community/lists/users/2015/09/27560.php 
 
>>> -- 
>>> M.Sc. Tobias Klöffel
>>> ===
>>> Interdisciplinary Center for Molecular Materials (ICMM)
>>> and Computer-Chemistry-Center (CCC)
>>> Department Chemie und Pharmazie
>>> Friedrich-Alexander-Universität

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Gilles Gouaillardet

on linux, you can look at /proc/self/status and search allowed_cpus_list
or you can use the sched_getaffinity system call

note that in some (hopefully rare)cases, this will return different results
than hwloc

On Sunday, September 13, 2015, Saliya Ekanayake  wrote:

> Thank you, I'll try this. Also, is there a way to know which core a
> process is bound to within the program other than executing something like
> taskset  from program?
>
> On Sun, Sep 13, 2015 at 10:05 AM, Ralph Castain  > wrote:
>
>> Actually, the error was correct - it was me that was incorrect. The
>> correct set of options would be:
>>
>> —map-by ppr:12_node —bind-to core —cpu-set=0,2,4,…
>>
>> Sorry about the confusion
>>
>>
>> On Sep 13, 2015, at 2:43 AM, Ralph Castain > > wrote:
>>
>> The rankfile will certainly do it, but that error is a bug and I’ll have
>> to fix it.
>>
>>
>> On Sep 13, 2015, at 1:10 AM, Saliya Ekanayake > > wrote:
>>
>> I could get it working by manually generating a rankfile all the ranks
>> and not using any --map-by options.
>>
>> I'll try the --map-by core as well
>>
>> On Sun, Sep 13, 2015 at 3:59 AM, Tobias Kloeffel > > wrote:
>>
>>> Hi,
>>> use: --map-by core
>>>
>>> regards,
>>> Tobias
>>>
>>>
>>> On 09/13/2015 09:41 AM, Saliya Ekanayake wrote:
>>>
>>> I tried,
>>>
>>>  --map-by ppr:12:node --slot-list 0,2,4,6,8,10,12,14,16,18,20,22
>>> --bind-to core -np 12
>>>
>>> but it complains,
>>>
>>> "Conflicting directives for binding policy are causing the policy
>>> to be redefined:
>>>
>>>   New policy:   socket
>>>   Prior policy:  CORE
>>>
>>> Please check that only one policy is defined.
>>> "
>>>
>>> On Sun, Sep 13, 2015 at 2:57 AM, Ralph Castain >> > wrote:
>>>
 Try something like this instead:

 —map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…

 You’ll have to play a bit with the core numbers in the slot-list to get
 the numbering right as I don’t know how your machine numbers them, and I
 can’t guarantee it will work - but it’s worth a shot. If it doesn’t, then I
 may have to add an option for such purposes

 Ralph

 On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake > wrote:

 Hi,

 We've a machine as in the following picture. I'd like to run 12 MPI
 procs per node each bound to 1 core, but like shown in blue dots in the
 pictures. I can use the following command to run 12 procs per node, but
 PE=1 makes all the 12 processes will run in just 1 socket. PE=2 will make a
 process bind to 2 cores, which is not what I want.

 --map-by ppr:12:node:PE=1,SPAN

 Thank you,
 Saliya

 

 --
 Saliya Ekanayake
 Ph.D. Candidate | Research Assistant
 School of Informatics and Computing | Digital Science Center
 Indiana University, Bloomington
 Cell 812-391-4914
 http://saliya.org
 ___
 users mailing list
 us...@open-mpi.org 
 Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
 Link to this post:
 http://www.open-mpi.org/community/lists/users/2015/09/27558.php



 ___
 users mailing list
 us...@open-mpi.org 
 Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
 Link to this post:
 http://www.open-mpi.org/community/lists/users/2015/09/27559.php

>>>
>>>
>>>
>>> --
>>> Saliya Ekanayake
>>> Ph.D. Candidate | Research Assistant
>>> School of Informatics and Computing | Digital Science Center
>>> Indiana University, Bloomington
>>> Cell 812-391-4914
>>> http://saliya.org
>>>
>>>
>>> ___
>>> users mailing listus...@open-mpi.org 
>>> 
>>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>>>
>>> Link to this post: 
>>> http://www.open-mpi.org/community/lists/users/2015/09/27560.php
>>>
>>>
>>> --
>>> M.Sc. Tobias Klöffel
>>> ===
>>> Interdisciplinary Center for Molecular Materials (ICMM)
>>> and Computer-Chemistry-Center (CCC)
>>> Department Chemie und Pharmazie
>>> Friedrich-Alexander-Universität Erlangen-Nürnberg
>>> Nägelsbachstr. 25
>>> D-91052 Erlangen, Germany
>>>
>>> Room: 2.307
>>> Phone: +49 (0) 9131 / 85 - 20421
>>> Fax: +49 (0) 9131 / 85 - 26565
>>>
>>>

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Saliya Ekanayake

Thank you, I'll try this. Also, is there a way to know which core a process
is bound to within the program other than executing something like taskset
 from program?

On Sun, Sep 13, 2015 at 10:05 AM, Ralph Castain  wrote:

> Actually, the error was correct - it was me that was incorrect. The
> correct set of options would be:
>
> —map-by ppr:12_node —bind-to core —cpu-set=0,2,4,…
>
> Sorry about the confusion
>
>
> On Sep 13, 2015, at 2:43 AM, Ralph Castain  wrote:
>
> The rankfile will certainly do it, but that error is a bug and I’ll have
> to fix it.
>
>
> On Sep 13, 2015, at 1:10 AM, Saliya Ekanayake  wrote:
>
> I could get it working by manually generating a rankfile all the ranks and
> not using any --map-by options.
>
> I'll try the --map-by core as well
>
> On Sun, Sep 13, 2015 at 3:59 AM, Tobias Kloeffel 
> wrote:
>
>> Hi,
>> use: --map-by core
>>
>> regards,
>> Tobias
>>
>>
>> On 09/13/2015 09:41 AM, Saliya Ekanayake wrote:
>>
>> I tried,
>>
>>  --map-by ppr:12:node --slot-list 0,2,4,6,8,10,12,14,16,18,20,22
>> --bind-to core -np 12
>>
>> but it complains,
>>
>> "Conflicting directives for binding policy are causing the policy
>> to be redefined:
>>
>>   New policy:   socket
>>   Prior policy:  CORE
>>
>> Please check that only one policy is defined.
>> "
>>
>> On Sun, Sep 13, 2015 at 2:57 AM, Ralph Castain  wrote:
>>
>>> Try something like this instead:
>>>
>>> —map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…
>>>
>>> You’ll have to play a bit with the core numbers in the slot-list to get
>>> the numbering right as I don’t know how your machine numbers them, and I
>>> can’t guarantee it will work - but it’s worth a shot. If it doesn’t, then I
>>> may have to add an option for such purposes
>>>
>>> Ralph
>>>
>>> On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake  wrote:
>>>
>>> Hi,
>>>
>>> We've a machine as in the following picture. I'd like to run 12 MPI
>>> procs per node each bound to 1 core, but like shown in blue dots in the
>>> pictures. I can use the following command to run 12 procs per node, but
>>> PE=1 makes all the 12 processes will run in just 1 socket. PE=2 will make a
>>> process bind to 2 cores, which is not what I want.
>>>
>>> --map-by ppr:12:node:PE=1,SPAN
>>>
>>> Thank you,
>>> Saliya
>>>
>>> 
>>>
>>> --
>>> Saliya Ekanayake
>>> Ph.D. Candidate | Research Assistant
>>> School of Informatics and Computing | Digital Science Center
>>> Indiana University, Bloomington
>>> Cell 812-391-4914
>>> http://saliya.org
>>> ___
>>> users mailing list
>>> us...@open-mpi.org
>>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>>> Link to this post:
>>> http://www.open-mpi.org/community/lists/users/2015/09/27558.php
>>>
>>>
>>>
>>> ___
>>> users mailing list
>>> us...@open-mpi.org
>>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>>> Link to this post:
>>> http://www.open-mpi.org/community/lists/users/2015/09/27559.php
>>>
>>
>>
>>
>> --
>> Saliya Ekanayake
>> Ph.D. Candidate | Research Assistant
>> School of Informatics and Computing | Digital Science Center
>> Indiana University, Bloomington
>> Cell 812-391-4914
>> http://saliya.org
>>
>>
>> ___
>> users mailing listus...@open-mpi.org
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>>
>> Link to this post: 
>> http://www.open-mpi.org/community/lists/users/2015/09/27560.php
>>
>>
>> --
>> M.Sc. Tobias Klöffel
>> ===
>> Interdisciplinary Center for Molecular Materials (ICMM)
>> and Computer-Chemistry-Center (CCC)
>> Department Chemie und Pharmazie
>> Friedrich-Alexander-Universität Erlangen-Nürnberg
>> Nägelsbachstr. 25
>> D-91052 Erlangen, Germany
>>
>> Room: 2.307
>> Phone: +49 (0) 9131 / 85 - 20421
>> Fax: +49 (0) 9131 / 85 - 26565
>>
>> ===
>>
>>
>> E-mail: tobias.kloef...@fau.de
>>
>>
>> ___
>> users mailing list
>> us...@open-mpi.org
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>> Link to this post:
>> http://www.open-mpi.org/community/lists/users/2015/09/27561.php
>>
>
>
>
> --
> Saliya Ekanayake
> Ph.D. Candidate | Research Assistant
> School of Informatics and Computing | Digital Science Center
> Indiana University, Bloomington
> Cell 812-391-4914
> http://saliya.org
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/09/27562.php
>
>
>
>
> ___
> users mailing list
> us...@open-mpi.org
> Subscription:

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Ralph Castain

Actually, the error was correct - it was me that was incorrect. The correct set 
of options would be:

—map-by ppr:12_node —bind-to core —cpu-set=0,2,4,…

Sorry about the confusion


> On Sep 13, 2015, at 2:43 AM, Ralph Castain  wrote:
> 
> The rankfile will certainly do it, but that error is a bug and I’ll have to 
> fix it.
> 
> 
>> On Sep 13, 2015, at 1:10 AM, Saliya Ekanayake > > wrote:
>> 
>> I could get it working by manually generating a rankfile all the ranks and 
>> not using any --map-by options.
>> 
>> I'll try the --map-by core as well
>> 
>> On Sun, Sep 13, 2015 at 3:59 AM, Tobias Kloeffel > > wrote:
>> Hi,
>> use: --map-by core
>> 
>> regards,
>> Tobias
>> 
>> 
>> On 09/13/2015 09:41 AM, Saliya Ekanayake wrote:
>>> I tried,
>>> 
>>>  --map-by ppr:12:node --slot-list 0,2,4,6,8,10,12,14,16,18,20,22 --bind-to 
>>> core -np 12
>>> 
>>> but it complains,
>>> 
>>> "Conflicting directives for binding policy are causing the policy
>>> to be redefined:
>>> 
>>>   New policy:   socket
>>>   Prior policy:  CORE
>>> 
>>> Please check that only one policy is defined.
>>> "
>>> 
>>> On Sun, Sep 13, 2015 at 2:57 AM, Ralph Castain >> > wrote:
>>> Try something like this instead:
>>> 
>>> —map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…
>>> 
>>> You’ll have to play a bit with the core numbers in the slot-list to get the 
>>> numbering right as I don’t know how your machine numbers them, and I can’t 
>>> guarantee it will work - but it’s worth a shot. If it doesn’t, then I may 
>>> have to add an option for such purposes
>>> 
>>> Ralph
>>> 
 On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake > wrote:
 
 Hi,
 
 We've a machine as in the following picture. I'd like to run 12 MPI procs 
 per node each bound to 1 core, but like shown in blue dots in the 
 pictures. I can use the following command to run 12 procs per node, but 
 PE=1 makes all the 12 processes will run in just 1 socket. PE=2 will make 
 a process bind to 2 cores, which is not what I want. 
 
 --map-by ppr:12:node:PE=1,SPAN
 
 Thank you,
 Saliya
 
 
 
 -- 
 Saliya Ekanayake
 Ph.D. Candidate | Research Assistant
 School of Informatics and Computing | Digital Science Center
 Indiana University, Bloomington
 Cell 812-391-4914 
  http://saliya.org 
 ___
 users mailing list
 us...@open-mpi.org 
 Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
 
 Link to this post: 
 http://www.open-mpi.org/community/lists/users/2015/09/27558.php 
 
>>> 
>>> ___
>>> users mailing list
>>> us...@open-mpi.org 
>>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
>>> 
>>> Link to this post: 
>>> http://www.open-mpi.org/community/lists/users/2015/09/27559.php 
>>> 
>>> 
>>> 
>>> 
>>> -- 
>>> Saliya Ekanayake
>>> Ph.D. Candidate | Research Assistant
>>> School of Informatics and Computing | Digital Science Center
>>> Indiana University, Bloomington
>>> Cell 812-391-4914 
>>> http://saliya.org 
>>> 
>>> ___
>>> users mailing list
>>> us...@open-mpi.org 
>>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
>>> 
>>> Link to this post: 
>>> http://www.open-mpi.org/community/lists/users/2015/09/27560.php 
>>> 
>> -- 
>> M.Sc. Tobias Klöffel
>> ===
>> Interdisciplinary Center for Molecular Materials (ICMM)
>> and Computer-Chemistry-Center (CCC)
>> Department Chemie und Pharmazie
>> Friedrich-Alexander-Universität Erlangen-Nürnberg
>> Nägelsbachstr. 25
>> D-91052 Erlangen, Germany
>> 
>> Room: 2.307
>> Phone: +49 (0) 9131 / 85 - 20421 
>> 
>> Fax: +49 (0) 9131 / 85 - 26565 
>> 
>> 
>> ===
>> 
>> 
>> E-mail: tobias.kloef...@fau.de 
>> ___
>> users mailing list
>> us...@open-mpi.org 
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
>>

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Ralph Castain

The rankfile will certainly do it, but that error is a bug and I’ll have to fix 
it.


> On Sep 13, 2015, at 1:10 AM, Saliya Ekanayake  wrote:
> 
> I could get it working by manually generating a rankfile all the ranks and 
> not using any --map-by options.
> 
> I'll try the --map-by core as well
> 
> On Sun, Sep 13, 2015 at 3:59 AM, Tobias Kloeffel  > wrote:
> Hi,
> use: --map-by core
> 
> regards,
> Tobias
> 
> 
> On 09/13/2015 09:41 AM, Saliya Ekanayake wrote:
>> I tried,
>> 
>>  --map-by ppr:12:node --slot-list 0,2,4,6,8,10,12,14,16,18,20,22 --bind-to 
>> core -np 12
>> 
>> but it complains,
>> 
>> "Conflicting directives for binding policy are causing the policy
>> to be redefined:
>> 
>>   New policy:   socket
>>   Prior policy:  CORE
>> 
>> Please check that only one policy is defined.
>> "
>> 
>> On Sun, Sep 13, 2015 at 2:57 AM, Ralph Castain > > wrote:
>> Try something like this instead:
>> 
>> —map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…
>> 
>> You’ll have to play a bit with the core numbers in the slot-list to get the 
>> numbering right as I don’t know how your machine numbers them, and I can’t 
>> guarantee it will work - but it’s worth a shot. If it doesn’t, then I may 
>> have to add an option for such purposes
>> 
>> Ralph
>> 
>>> On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake >> > wrote:
>>> 
>>> Hi,
>>> 
>>> We've a machine as in the following picture. I'd like to run 12 MPI procs 
>>> per node each bound to 1 core, but like shown in blue dots in the pictures. 
>>> I can use the following command to run 12 procs per node, but PE=1 makes 
>>> all the 12 processes will run in just 1 socket. PE=2 will make a process 
>>> bind to 2 cores, which is not what I want. 
>>> 
>>> --map-by ppr:12:node:PE=1,SPAN
>>> 
>>> Thank you,
>>> Saliya
>>> 
>>> 
>>> 
>>> -- 
>>> Saliya Ekanayake
>>> Ph.D. Candidate | Research Assistant
>>> School of Informatics and Computing | Digital Science Center
>>> Indiana University, Bloomington
>>> Cell 812-391-4914 
>>>  http://saliya.org 
>>> ___
>>> users mailing list
>>> us...@open-mpi.org 
>>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
>>> 
>>> Link to this post: 
>>> http://www.open-mpi.org/community/lists/users/2015/09/27558.php 
>>> 
>> 
>> ___
>> users mailing list
>> us...@open-mpi.org 
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
>> 
>> Link to this post: 
>> http://www.open-mpi.org/community/lists/users/2015/09/27559.php 
>> 
>> 
>> 
>> 
>> -- 
>> Saliya Ekanayake
>> Ph.D. Candidate | Research Assistant
>> School of Informatics and Computing | Digital Science Center
>> Indiana University, Bloomington
>> Cell 812-391-4914 
>> http://saliya.org 
>> 
>> ___
>> users mailing list
>> us...@open-mpi.org 
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
>> 
>> Link to this post: 
>> http://www.open-mpi.org/community/lists/users/2015/09/27560.php 
>> 
> -- 
> M.Sc. Tobias Klöffel
> ===
> Interdisciplinary Center for Molecular Materials (ICMM)
> and Computer-Chemistry-Center (CCC)
> Department Chemie und Pharmazie
> Friedrich-Alexander-Universität Erlangen-Nürnberg
> Nägelsbachstr. 25
> D-91052 Erlangen, Germany
> 
> Room: 2.307
> Phone: +49 (0) 9131 / 85 - 20421 
> 
> Fax: +49 (0) 9131 / 85 - 26565 
> 
> 
> ===
> 
> 
> E-mail: tobias.kloef...@fau.de 
> ___
> users mailing list
> us...@open-mpi.org 
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users 
> 
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/09/27561.php 
> 
> 
> 
> 
> -- 
> Saliya Ekanayake
> Ph.D. Candidate | Research Assistant
> School of Informatics and Computing | Digital Science Center
> Indiana University, Bloomington
> Cell 812-391-4914
> http://saliya.org 
>

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Saliya Ekanayake

I could get it working by manually generating a rankfile all the ranks and
not using any --map-by options.

I'll try the --map-by core as well

On Sun, Sep 13, 2015 at 3:59 AM, Tobias Kloeffel 
wrote:

> Hi,
> use: --map-by core
>
> regards,
> Tobias
>
>
> On 09/13/2015 09:41 AM, Saliya Ekanayake wrote:
>
> I tried,
>
>  --map-by ppr:12:node --slot-list 0,2,4,6,8,10,12,14,16,18,20,22 --bind-to
> core -np 12
>
> but it complains,
>
> "Conflicting directives for binding policy are causing the policy
> to be redefined:
>
>   New policy:   socket
>   Prior policy:  CORE
>
> Please check that only one policy is defined.
> "
>
> On Sun, Sep 13, 2015 at 2:57 AM, Ralph Castain  wrote:
>
>> Try something like this instead:
>>
>> —map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…
>>
>> You’ll have to play a bit with the core numbers in the slot-list to get
>> the numbering right as I don’t know how your machine numbers them, and I
>> can’t guarantee it will work - but it’s worth a shot. If it doesn’t, then I
>> may have to add an option for such purposes
>>
>> Ralph
>>
>> On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake  wrote:
>>
>> Hi,
>>
>> We've a machine as in the following picture. I'd like to run 12 MPI procs
>> per node each bound to 1 core, but like shown in blue dots in the pictures.
>> I can use the following command to run 12 procs per node, but PE=1 makes
>> all the 12 processes will run in just 1 socket. PE=2 will make a process
>> bind to 2 cores, which is not what I want.
>>
>> --map-by ppr:12:node:PE=1,SPAN
>>
>> Thank you,
>> Saliya
>>
>> 
>>
>> --
>> Saliya Ekanayake
>> Ph.D. Candidate | Research Assistant
>> School of Informatics and Computing | Digital Science Center
>> Indiana University, Bloomington
>> Cell 812-391-4914
>> http://saliya.org
>> ___
>> users mailing list
>> us...@open-mpi.org
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>> Link to this post:
>> http://www.open-mpi.org/community/lists/users/2015/09/27558.php
>>
>>
>>
>> ___
>> users mailing list
>> us...@open-mpi.org
>> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>> Link to this post:
>> http://www.open-mpi.org/community/lists/users/2015/09/27559.php
>>
>
>
>
> --
> Saliya Ekanayake
> Ph.D. Candidate | Research Assistant
> School of Informatics and Computing | Digital Science Center
> Indiana University, Bloomington
> Cell 812-391-4914
> http://saliya.org
>
>
> ___
> users mailing listus...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
>
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/09/27560.php
>
>
> --
> M.Sc. Tobias Klöffel
> ===
> Interdisciplinary Center for Molecular Materials (ICMM)
> and Computer-Chemistry-Center (CCC)
> Department Chemie und Pharmazie
> Friedrich-Alexander-Universität Erlangen-Nürnberg
> Nägelsbachstr. 25
> D-91052 Erlangen, Germany
>
> Room: 2.307
> Phone: +49 (0) 9131 / 85 - 20421
> Fax: +49 (0) 9131 / 85 - 26565
>
> ===
>
>
> E-mail: tobias.kloef...@fau.de
>
>
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/09/27561.php
>



-- 
Saliya Ekanayake
Ph.D. Candidate | Research Assistant
School of Informatics and Computing | Digital Science Center
Indiana University, Bloomington
Cell 812-391-4914
http://saliya.org

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Tobias Kloeffel


Hi,
use: --map-by core

regards,
Tobias

On 09/13/2015 09:41 AM, Saliya Ekanayake wrote:

I tried,

 --map-by ppr:12:node --slot-list 0,2,4,6,8,10,12,14,16,18,20,22 
--bind-to core -np 12


but it complains,

"Conflicting directives for binding policy are causing the policy
to be redefined:

  New policy:   socket
  Prior policy:  CORE

Please check that only one policy is defined.
"

On Sun, Sep 13, 2015 at 2:57 AM, Ralph Castain > wrote:


Try something like this instead:

—map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…

You’ll have to play a bit with the core numbers in the slot-list
to get the numbering right as I don’t know how your machine
numbers them, and I can’t guarantee it will work - but it’s worth
a shot. If it doesn’t, then I may have to add an option for such
purposes

Ralph


On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake > wrote:

Hi,

We've a machine as in the following picture. I'd like to run 12
MPI procs per node each bound to 1 core, but like shown in blue
dots in the pictures. I can use the following command to run 12
procs per node, but PE=1 makes all the 12 processes will run in
just 1 socket. PE=2 will make a process bind to 2 cores, which is
not what I want.

--map-by ppr:12:node:PE=1,SPAN

Thank you,
Saliya



-- 
Saliya Ekanayake

Ph.D. Candidate | Research Assistant
School of Informatics and Computing | Digital Science Center
Indiana University, Bloomington
Cell 812-391-4914 
http://saliya.org 
___
users mailing list
us...@open-mpi.org 
Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
Link to this post:
http://www.open-mpi.org/community/lists/users/2015/09/27558.php



___
users mailing list
us...@open-mpi.org 
Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
Link to this post:
http://www.open-mpi.org/community/lists/users/2015/09/27559.php




--
Saliya Ekanayake
Ph.D. Candidate | Research Assistant
School of Informatics and Computing | Digital Science Center
Indiana University, Bloomington
Cell 812-391-4914
http://saliya.org


___
users mailing list
us...@open-mpi.org
Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
Link to this post: 
http://www.open-mpi.org/community/lists/users/2015/09/27560.php


--
M.Sc. Tobias Klöffel
===
Interdisciplinary Center for Molecular Materials (ICMM)
and Computer-Chemistry-Center (CCC)
Department Chemie und Pharmazie
Friedrich-Alexander-Universität Erlangen-Nürnberg
Nägelsbachstr. 25
D-91052 Erlangen, Germany

Room: 2.307
Phone: +49 (0) 9131 / 85 - 20421
Fax: +49 (0) 9131 / 85 - 26565

===


E-mail: tobias.kloef...@fau.de

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Saliya Ekanayake

I tried,

 --map-by ppr:12:node --slot-list 0,2,4,6,8,10,12,14,16,18,20,22 --bind-to
core -np 12

but it complains,

"Conflicting directives for binding policy are causing the policy
to be redefined:

  New policy:   socket
  Prior policy:  CORE

Please check that only one policy is defined.
"

On Sun, Sep 13, 2015 at 2:57 AM, Ralph Castain  wrote:

> Try something like this instead:
>
> —map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…
>
> You’ll have to play a bit with the core numbers in the slot-list to get
> the numbering right as I don’t know how your machine numbers them, and I
> can’t guarantee it will work - but it’s worth a shot. If it doesn’t, then I
> may have to add an option for such purposes
>
> Ralph
>
> On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake  wrote:
>
> Hi,
>
> We've a machine as in the following picture. I'd like to run 12 MPI procs
> per node each bound to 1 core, but like shown in blue dots in the pictures.
> I can use the following command to run 12 procs per node, but PE=1 makes
> all the 12 processes will run in just 1 socket. PE=2 will make a process
> bind to 2 cores, which is not what I want.
>
> --map-by ppr:12:node:PE=1,SPAN
>
> Thank you,
> Saliya
>
> 
>
> --
> Saliya Ekanayake
> Ph.D. Candidate | Research Assistant
> School of Informatics and Computing | Digital Science Center
> Indiana University, Bloomington
> Cell 812-391-4914
> http://saliya.org
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/09/27558.php
>
>
>
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post:
> http://www.open-mpi.org/community/lists/users/2015/09/27559.php
>



-- 
Saliya Ekanayake
Ph.D. Candidate | Research Assistant
School of Informatics and Computing | Digital Science Center
Indiana University, Bloomington
Cell 812-391-4914
http://saliya.org

Re: [OMPI users] Help with Specific Binding

2015-09-13 Thread Ralph Castain

Try something like this instead:

—map-by ppr:12:node —bind-to core —slot-list=0,2,4,6,8,…

You’ll have to play a bit with the core numbers in the slot-list to get the 
numbering right as I don’t know how your machine numbers them, and I can’t 
guarantee it will work - but it’s worth a shot. If it doesn’t, then I may have 
to add an option for such purposes

Ralph

> On Sep 12, 2015, at 7:39 PM, Saliya Ekanayake  wrote:
> 
> Hi,
> 
> We've a machine as in the following picture. I'd like to run 12 MPI procs per 
> node each bound to 1 core, but like shown in blue dots in the pictures. I can 
> use the following command to run 12 procs per node, but PE=1 makes all the 12 
> processes will run in just 1 socket. PE=2 will make a process bind to 2 
> cores, which is not what I want. 
> 
> --map-by ppr:12:node:PE=1,SPAN
> 
> Thank you,
> Saliya
> 
> 
> 
> -- 
> Saliya Ekanayake
> Ph.D. Candidate | Research Assistant
> School of Informatics and Computing | Digital Science Center
> Indiana University, Bloomington
> Cell 812-391-4914
> http://saliya.org 
> ___
> users mailing list
> us...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users
> Link to this post: 
> http://www.open-mpi.org/community/lists/users/2015/09/27558.php

[OMPI users] Help with Specific Binding

2015-09-12 Thread Saliya Ekanayake

Hi,

We've a machine as in the following picture. I'd like to run 12 MPI procs
per node each bound to 1 core, but like shown in blue dots in the pictures.
I can use the following command to run 12 procs per node, but PE=1 makes
all the 12 processes will run in just 1 socket. PE=2 will make a process
bind to 2 cores, which is not what I want.

--map-by ppr:12:node:PE=1,SPAN

Thank you,
Saliya

[image: Inline image 1]

-- 
Saliya Ekanayake
Ph.D. Candidate | Research Assistant
School of Informatics and Computing | Digital Science Center
Indiana University, Bloomington
Cell 812-391-4914
http://saliya.org

Re: [OMPI users] Help with Specific Binding

Re: [OMPI users] Help with Specific Binding

Re: [OMPI users] Help with Specific Binding

Re: [OMPI users] Help with Specific Binding

Re: [OMPI users] Help with Specific Binding

Re: [OMPI users] Help with Specific Binding

Re: [OMPI users] Help with Specific Binding

Re: [OMPI users] Help with Specific Binding

Re: [OMPI users] Help with Specific Binding

[OMPI users] Help with Specific Binding

10 matches

Site Navigation

Mail list logo

Footer information