[mpich-discuss] regarding parallel proramming error

Wesley Bland wbland at mcs.anl.gov
Fri Aug 23 09:26:58 CDT 2013


Re-adding MPICH-discuss. Please don't remove the mailing list from the CC.

I'm not sure what the exact problem is. It's not a problem with your code,
so it seems to be a problem with the setup. It could be that you have
multiple MPI installations on your machine. When you do `which mpirun`, is
it the same installation as `which mpicc` or `which mpiexec`? Do you have
the same versions of MPI installed on your remote machines?


On Fri, Aug 23, 2013 at 9:16 AM, urmila pol <urmilec at gmail.com> wrote:

> hello sir,
>
> I use mpich-3.0.4 version. What is the exact problem. I want
>  to know .
>
> With regards,
>
> Dr.Urmila R. Pol
> Assistant Professor,
> Department of computer Science,
> Shivaji University, Kolhapur.
>
>
> On Fri, Aug 23, 2013 at 5:56 AM, Wesley Bland <wbland at mcs.anl.gov> wrote:
>
>> You code looks pretty much fine to me. The only issue I see is that you
>> should pass argc and argv to MPI_Init, not NULL. That probably isn't
>> causing the issue though.
>>
>> Which version of MPICH are you using (mpicc -v)?
>>
>>
>>
>> On Fri, Aug 23, 2013 at 1:35 AM, urmila pol <urmilec at gmail.com> wrote:
>>
>>> Dear sir,
>>>
>>>      I have attached all the files and my errors into errorfile please
>>> guide me what happen
>>> in that case. program is not working for multiple host.
>>> my host file is mfile. code is hello2.c
>>> waitibg for ur reply.
>>>
>>> On 8/22/13, Wesley Bland <wbland at mcs.anl.gov> wrote:
>>> > Edit:
>>> >
>>> > Can you attach you machine file and the *smallest amount of code that
>>> will
>>> > reproduce the error*. It's tough to dig through huge amounts of someone
>>> > else's code to find the error and while you're trying to reduce the
>>> amount
>>> > of code, you often find the issue.
>>> >
>>> > Thanks,
>>> > Wesley
>>> >
>>> >
>>> > On Thu, Aug 22, 2013 at 3:46 AM, Mahesh Doijade
>>> > <maheshdoijade at gmail.com>wrote:
>>> >
>>> >> Can you attached your machinefile, your code and detailed error
>>> output.
>>> >>
>>> >>
>>> >> On Tue, Aug 20, 2013 at 4:11 PM, urmila pol <urmilec at gmail.com>
>>> wrote:
>>> >>
>>> >>> Dear sir,
>>> >>>
>>> >>> I was given following command it will give floating point error .
>>> >>> #mpirun -f mfile -np 2  hello4
>>> >>> Floating point exception
>>> >>>
>>> >>> if I give following it will execute for single node
>>> >>>
>>> >>> [root at masternode ~]# mpirun -np 2 hello4
>>> >>> Hello world from processor masternode.csd.com, rank 0 out of 2
>>> >>> processors
>>> >>> Hello world from processor masternode.csd.com, rank 1 out of 2
>>> >>> processors
>>> >>> [root at masternode ~]#
>>> >>>
>>> >>> Please guide me for above error .waiting for ur reply.
>>> >>>
>>> >>>
>>> >>>
>>> >>>
>>> >>>
>>> >>>
>>> >>>
>>> >>> On 8/20/13, urmila pol <urmilec at gmail.com> wrote:
>>> >>> > I do everything now i have one error flaoting point exception i am
>>> >>> > running my program now
>>> >>> > waiting for ur reply
>>> >>>
>>> >>> >
>>> >>> > With regards,
>>> >>> >
>>> >>> > Dr.Urmila R. Pol
>>> >>> > Assistant Professor,
>>> >>> > Department of computer Science,
>>> >>> > Shivaji University, Kolhapur.
>>> >>> >
>>> >>> >
>>> >>> > On Tue, Aug 20, 2013 at 10:44 AM, Mahesh Doijade
>>> >>> > <maheshdoijade at gmail.com>wrote:
>>> >>>
>>> >>> >
>>> >>> >>
>>> >>> >>      The issue seems to be that you have probably not created a
>>> >>> >> machinefile which consists of node name to number of processes
>>> >>> >> mapping
>>> >>> >> and
>>> >>> >> this file should be present in the same dir. of the other nodes on
>>> >>> which
>>> >>> >> you are launching other MPI processes, this is required in case
>>> you
>>> >>> want
>>> >>> >> to
>>> >>> >> run the MPI program across the two or more nodes, you can find
>>> more
>>> >>> >> details
>>> >>> >> of the format in which this information has to be in the
>>> machinefile
>>> >>> and
>>> >>> >> the other requisite steps need for running your MPI program across
>>> >>> >> multiple
>>> >>> >> nodes from this link
>>> >>> >>
>>> http://wiki.mpich.org/mpich/index.php/Using_the_Hydra_Process_Manager
>>> >>> >>        Also, in order to launch the processes across multiple
>>> nodes
>>> >>> >> you
>>> >>> >> need to do ssh passwordless login among those nodes, in case you
>>> are
>>> >>> >> using
>>> >>> >> ssh as the connection mechanism, you can find more details about
>>> the
>>> >>> >> procedure for ssh passwordless login over here
>>> >>> >> http://www.techlitic.com/ssh-passwordless-login.html
>>> >>> >>
>>> >>> >>
>>> >>> >> Regards,
>>> >>> >> Mahesh Doijade,
>>> >>> >> http://www.techdarting.com/
>>> >>> >>
>>> >>> >>
>>> >>> >> On Tue, Aug 20, 2013 at 8:34 AM, Pavan Balaji <balaji at mcs.anl.gov
>>> >
>>> >>> wrote:
>>> >>> >>
>>> >>> >>>
>>> >>> >>> I'm pasting the error inline, for others' information.
>>> >>> >>>
>>> >>> >>> FYI, the error is "unable to open host file: machinefile".
>>> >>> >>>
>>> >>> >>> ----8<----
>>> >>> >>> mpirun -f machinefile -n 2 /home/urp/mpich-install/hello
>>> >>> >>> [mpiexec at masternode.csd.com] HYDU_parse_hostfile
>>> >>> >>> (./utils/args/args.c:323): unable to open host file: machinefile
>>> >>> >>> [mpiexec at masternode.csd.com] mfile_fn (./ui/mpich/utils.c:341):
>>> >>> >>> error
>>> >>> >>> parsing hostfile
>>> >>> >>> [mpiexec at masternode.csd.com] match_arg
>>> (./utils/args/args.c:153):
>>> >>> match
>>> >>> >>> handler returned error
>>> >>> >>> [mpiexec at masternode.csd.com] HYDU_parse_array
>>> >>> (./utils/args/args.c:175):
>>> >>> >>> argument matching returned error
>>> >>> >>> [mpiexec at masternode.csd.com] parse_args
>>> (./ui/mpich/utils.c:1609):
>>> >>> error
>>> >>> >>> parsing input array
>>> >>> >>> [mpiexec at masternode.csd.com] HYD_uii_mpx_get_parameters
>>> >>> >>> (./ui/mpich/utils.c:1660): unable to parse user arguments
>>> >>> >>> [mpiexec at masternode.csd.com] main (./ui/mpich/mpiexec.c:153):
>>> error
>>> >>> >>> parsing parameters
>>> >>> >>> [root at masternode ~]#
>>> >>> >>> ----8<----
>>> >>> >>>
>>> >>> >>> On 08/19/2013 01:27 AM, urmila pol wrote:
>>> >>> >>>
>>> >>> >>>> Dear sir,
>>> >>> >>>>
>>> >>> >>>> I have attaching the screenshots of my errors please guide me
>>> >>> .waiting
>>> >>> >>>> for reply.
>>> >>> >>>>
>>> >>> >>>>
>>> >>> >>>> With regards,
>>> >>> >>>>
>>> >>> >>>> Dr.Urmila R. Pol
>>> >>> >>>> Assistant Professor,
>>> >>> >>>> Department of computer Science,
>>> >>> >>>> Shivaji University, Kolhapur.
>>> >>> >>>>
>>> >>> >>>>
>>> >>> >>> --
>>> >>> >>> Pavan Balaji
>>> >>> >>> http://www.mcs.anl.gov/~balaji
>>> >>> >>> ______________________________**_________________
>>> >>>
>>> >>> >>> discuss mailing list     discuss at mpich.org
>>> >>> >>> To manage subscription options or unsubscribe:
>>> >>> >>> https://lists.mpich.org/**mailman/listinfo/discuss<
>>> >>> https://lists.mpich.org/mailman/listinfo/discuss>
>>> >>> >>>
>>> >>> >>
>>> >>> >>
>>> >>> >>
>>> >>> >>
>>> >>> >>
>>> >>> >>
>>> >>> >>
>>> >>> >>
>>> >>> >
>>> >>>
>>> >>>
>>> >>> --
>>> >>>
>>> >>> With regards,
>>> >>>
>>> >>> Dr.Urmila R. Pol
>>> >>> Assistant Professor,
>>> >>> Department of computer Science,
>>> >>> Shivaji University, Kolhapur.
>>> >>>
>>> >>
>>> >>
>>> >>
>>> >> --
>>> >>
>>> >> Regards,
>>> >> -- Mahesh Doijade
>>> >>
>>> >>
>>> >>
>>> >>
>>> >>
>>> >> _______________________________________________
>>> >> discuss mailing list     discuss at mpich.org
>>> >> To manage subscription options or unsubscribe:
>>> >> https://lists.mpich.org/mailman/listinfo/discuss
>>> >>
>>> >
>>>
>>>
>>> --
>>> With regards,
>>>
>>> Dr.Urmila R. Pol
>>> Assistant Professor,
>>> Department of computer Science,
>>> Shivaji University, Kolhapur.
>>>
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mpich.org/pipermail/discuss/attachments/20130823/f9118790/attachment.html>


More information about the discuss mailing list