As Robert mentioned, you might want to make sure that the LAM
installation you're using actually has the TM boot module installed.
Check the output from the laminfo command, for example.
Another reason that the tm module may not be found is that the tm
module may have elected not to run (the error message in this case
would be a little misleading; sorry). If you're not running in a PBS
job, then the tm module will automatically exclude itself from the
possibility of selection at run time. I assume that you *are*
running in a PBS job (you implied as much in your text); I just
mention this fact for completeness.
Finally, note that you should not need the "-ssi boot tm" switches to
lamboot when running inside a PBS job. LAM should automatically
detect that it is running in a PBS job and do the Right Things.
Indeed, if you are running in a PBS job, LAM will ignore any
specified hostfile and only use the data provided by PBS.
On Jun 20, 2007, at 8:32 PM, SCIPIONI Roberto wrote:
> As far as I know with the tm module
> lamboot shoud be enough becuase the tm module is found
> automatically but make sure that the module tm is in the standard
> path otherwise you may want to use
>
> --with-boot-tm=PATH
>
> in the LAM configure
>
> Cheers
>
>
> Roberto
>
> ----- Original Message -----
> From: "Taylor, Trent Matthew (UMR-Student)" <ttaylor_at_[hidden]>
> Date: Wednesday, June 20, 2007 8:10 am
> Subject: LAM: The boot module named "tm" could not be found
>
>> I am running a parallel program with a PBS scheduler. I have ran
>> the same code on a different system and used the following command
>> in my job script:
>>
>> lamboot -v -ssi boot tm
>>
>> and it worked on another system. But this new one has the
>> following in the standard error file:
>>
>> -------------------------------------------------------------------
>> ----------
>> The boot module named "tm" could not be found.
>> This typically means that you misspelled the desired module name,
>> used
>> the wrong name entirely, or the module has decided that it does not
>> want to run in this environment.
>> -------------------------------------------------------------------
>> ----------
>> Lamnodes Failed!
>> Check if you had booted lam before calling mpiexec else use -
>> machinefile to pass host file to mpiexec
>>
>>
>>
>>
>> I have also tried using:
>>
>> lamboot $PBS_NODEFILE
>>
>> but I get the following error in my output stream:
>>
>> Missing: program name
>> Program lam_appschema_Et0oze either does not exist, is not
>> executable, or is an erroneous argument to mpirun.
>>
>> Any ideas how I can get lamboot to find the tm module? Thanks.
>>
>> _______________________________________________
>> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
--
Jeff Squyres
Cisco Systems
|
