Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |  

This web mail archive is frozen.

This page is part of a frozen web archive of this mailing list.

You can still navigate around this archive, but know that no new mails have been added to it since July of 2016.

Click here to be taken to the new web archives of this list; it includes all the mails that are in this frozen archive plus all new mails that have been sent to the list since it was migrated to the new archives.

Subject: Re: [OMPI devel] 1.3 runtime message ?
From: Jim Langston (Jim.Langston_at_[hidden])
Date: 2009-02-20 09:51:11


Don Kerr wrote:
>
>
> On 02/20/09 09:18, Jim Langston wrote:
>> Jeff Squyres wrote:
>>> On Feb 19, 2009, at 4:30 PM, Jim Langston wrote:
>>>
>>>> I'm just wondering if I compiled incorrectly, or if the message is
>>>> generated because uDAPL is the default
>>>> interface (I'm on OpenSolaris), which I'm wondering if I can stop
>>>> from being looked at, either by
>>>> compiling differently, or from the command line.
>>>>
>>>> This doesn't stop the message either, but stops the noisy stuff:
>>>>
>>>> orterun --mca orte_base_help_aggregate 0 --mca blt tcp -np 2 hello_c
>>>
>>>
>>> I can't quite parse your statement "doesn't stop the message" vs.
>>> "but stops the noisy stuff". Can you clarify what each of those means?
>>>
>>> Running with "--mca btl tcp,sm,self" should prevent OMPI v1.3 from
>>> opening the udapl BTL, and therefore that error message shouldn't
>>> appear.
>> I'm not considering this an error, the job runs without any problems.
>> The orte_base_help_aggregate set to 0 stops some of the other
>> messages that occur, which is great, I appreciate that option. But,
>> the option does not stop the point-to-point message,
>> which I'm assuming, means that something is happening underneath the
>> sheets and I'm being told it is getting done. With your
>> option:
>>
>> langston_at_alpha:~/OPENMPI/openmpi-1.3/examples$ orterun --mca
>> orte_base_help_aggregate 0 --mca blt tcp,sm,self -np 2 hello_c
> If this is the actual run command line I believe there is a typo:
> "--mca blt tcp,sm,self" should be "btl"
> which should eliminate the point-to-point message.

yes, definitely a typo - geeezzzz !! message gone, both --mca btl
tcp,self and --mca btl tcp,sm,self do just fine.

langston_at_alpha:~/OPENMPI/openmpi-1.3/examples$ orterun --mca btl
tcp,self -np 2 hello_c
Hello, world, I am 1 of 2
Hello, world, I am 0 of 2

and

langston_at_alpha:~/OPENMPI/openmpi-1.3/examples$ orterun --mca btl
tcp,sm,self -np 2 hello_c
Hello, world, I am 0 of 2
Hello, world, I am 1 of 2

>> --------------------------------------------------------------------------
>>
>> [[56072,1],0]: A high-performance Open MPI point-to-point messaging
>> module
>> was unable to find any relevant network interfaces:
>>
>> Module: uDAPL
>> Host: alpha
>>
>> Another transport will be used instead, although this may result in
>> lower performance.
>> --------------------------------------------------------------------------
>>
>> --------------------------------------------------------------------------
>>
>> [[56072,1],1]: A high-performance Open MPI point-to-point messaging
>> module
>> was unable to find any relevant network interfaces:
>>
>> Module: uDAPL
>> Host: alpha
>>
>> Another transport will be used instead, although this may result in
>> lower performance.
>> --------------------------------------------------------------------------
>>
>> Hello, world, I am 0 of 2
>> Hello, world, I am 1 of 2
>>
>> ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
>>
>>
>> Without orte_base_help_aggregate 0:
>>
>> langston_at_alpha:~/OPENMPI/openmpi-1.3/examples$ orterun --mca blt
>> tcp,sm,self -np 2 hello_c
>> --------------------------------------------------------------------------
>>
>> [[56119,1],0]: A high-performance Open MPI point-to-point messaging
>> module
>> was unable to find any relevant network interfaces:
>>
>> Module: uDAPL
>> Host: alpha
>>
>> Another transport will be used instead, although this may result in
>> lower performance.
>> --------------------------------------------------------------------------
>>
>> Hello, world, I am 0 of 2
>> Hello, world, I am 1 of 2
>> [alpha:10021] 1 more process has sent help message
>> help-mpi-btl-base.txt / btl:no-nics
>> [alpha:10021] Set MCA parameter "orte_base_help_aggregate" to 0 to
>> see all help / error messages
>>
>>
>> Thanks,
>>
>> Jim
>>
> _______________________________________________
> devel mailing list
> devel_at_[hidden]
> http://www.open-mpi.org/mailman/listinfo.cgi/devel

-- 
/////////////////////////////////////////////
Jim Langston
Sun Microsystems, Inc.
(877) 854-5583 (AccessLine)
(513) 702-4741 (Cell)
AIM: jl9594
jim.langston_at_[hidden]