- discuss - mailman3.cels.anl.gov

Re: [mpich-discuss] Hydra torque support issues
by Pavan Balaji 19 Apr '13

19 Apr '13

On 04/18/2013 06:56 PM US Central Time, Bharath Ramesh wrote: >> I'm particularly interested in whether mpiexec recognizes pbs as a >> possible launcher or not. If it recognizes pbs as a launcher, it should >> automatically select it where possible. > > It is using pbs as the launcher. The problem is with the setting of the > session id for the forked hydra_pmi_proxy process on mother superior. > Since the hydra_pmi_proxy process has the session id set to the bash > process instead mpi_exec process. As a result of this torque is unable > to track the resources consumed by the process. If you look at the > openmpi processes their session id is set correctly and torque is able > to track the resources consumed. Ah, OK. Sorry I lost track of the original issue in this email chain. You are right about this issue. The basic problem is that launching proxies using tm_spawn tends to create a different session ID for them. I've created a ticket to track it: https://trac.mpich.org/projects/mpich/ticket/1812 -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] seeking help
by Pavan Balaji 19 Apr '13

19 Apr '13

What problem are you facing? -- Pavan On 04/19/2013 12:44 AM US Central Time, manzoor Ahmed wrote: > HI > > I am doing Parallel Discrete event simulation based on multicore > computers using optimistic synchronization and message passing Interface > (MPI) libraray along with C++ . > > I am facing problem to interface C++ & MPI. > Please guide me to sort out my problem. > > M.G.Ahmed > > > Manzoor G. Ahmed > I/C HOD > ELectronics Department > mob:9881641949 > Dr. N.P.Hirani Institue of Polytechnic, > Pusad ,dist. Yavatmal (M.S,) > <http://sigads.rediff.com/RealMedia/ads/click_nx.ads/www.rediffmail.com/sign…> > > > Get your own *FREE* website and domain with business email solutions, > click here > <http://track.rediff.com/click?url=___http://hosting.rediff.com/rediffmailpr…> > > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

seeking help
by manzoor Ahmed 19 Apr '13

19 Apr '13

HI I am doing Parallel Discrete event simulation based on multicore computers using optimistic synchronization and message passing Interface (MPI) libraray along with C++ . I am facing problem to interface C++ & MPI. Please guide me to sort out my problem. M.G.Ahmed Manzoor G. Ahmed I/C HOD ELectronics Department mob:9881641949 Dr. N.P.Hirani Institue of Polytechnic, Pusad ,dist. Yavatmal (M.S,)

1 0

Re: [mpich-discuss] mpich - windows
by Dave Goodell 18 Apr '13

18 Apr '13

On Apr 18, 2013, at 9:47 PM CDT, Richard Muddle <rlmuddle(a)googlemail.com> wrote: > Windows installers for mpich haven't been release since 2.1.4.1p1 - is Windows supported for newer versions? Not at this time. The likelihood of support returning any time soon seems slim. Rather, it's more likely that we will further strip out Windows-specific code as that support bit rots. > Apart from Microsoft and Intel, are there any other alternatives for Windows? I think some versions of Open MPI have supported Windows in the past, though IIUC they are facing similar problems maintaining their Windows support at this time. Lack of Windows expertise, funding, and general development bandwidth make supporting Windows challenging for both projects. -Dave

1 0

Re: [mpich-discuss] mpich2.1.4.1p1: intercomm_create.c
by Dave Goodell 18 Apr '13

18 Apr '13

On Apr 18, 2013, at 9:43 PM CDT, Richard Muddle <rlmuddle(a)googlemail.com> wrote: > In src/mpi/comm/intercomm_create.c from lines 58, the calculation of maxi seems to be incorrect when maxlpid is a multiple of 32. > > /* Compute the max index and zero the pids array */ > maxi = (maxlpid + 31) / 32; > > if (maxi >= MAX_LPID32_ARRAY) { > MPIU_CHKLMEM_MALLOC(lpidmask,int32_t*,maxi*sizeof(int32_t), > mpi_errno,"lpidmask"); > } Hi Richard, Thanks for letting us know. I believe I already fixed this bug a few years ago: http://trac.mpich.org/projects/mpich/ticket/1574 http://git.mpich.org/mpich.git/commitdiff/40d88ed0 http://git.mpich.org/mpich.git/commitdiff/7159a5f9 Looks like those fixes were released in MPICH2 1.5b2: ----8<---- % git describe --contains 7159a5f9 v1.5b2~192 ----8<---- -Dave

1 0

mpich - windows
by Richard Muddle 18 Apr '13

18 Apr '13

Hi, Windows installers for mpich haven't been release since 2.1.4.1p1 - is Windows supported for newer versions? Apart from Microsoft and Intel, are there any other alternatives for Windows? Thank you, Richard

1 0

mpich2.1.4.1p1: intercomm_create.c
by Richard Muddle 18 Apr '13

18 Apr '13

In src/mpi/comm/intercomm_create.c from lines 58, the calculation of maxi seems to be incorrect when maxlpid is a multiple of 32. /* Compute the max index and zero the pids array */ maxi = (maxlpid + 31) / 32; if (maxi >= MAX_LPID32_ARRAY) { MPIU_CHKLMEM_MALLOC(lpidmask,int32_t*,maxi*sizeof(int32_t), mpi_errno,"lpidmask"); } else { lpidmask = lpidmaskPrealloc; } for (i=0; i<maxi; i++) lpidmask[i] = 0;

1 0

Re: [mpich-discuss] Hydra torque support issues
by Bharath Ramesh 18 Apr '13

18 Apr '13

On 4/18/2013 7:04 PM, Pavan Balaji wrote: > Bharath, > > Can you send the output of "mpiexec -info"? mpiexec -info HYDRA build details: Version: 3.0.2 Release Date: Wed Feb 27 19:12:19 EST 2013 CC: gcc -Wl,-rpath,/opt/apps/gcc/4.5.3/lib64 CXX: g++ -Wl,-rpath,/opt/apps/gcc/4.5.3/lib64 F77: gfortran -Wl,-rpath,/opt/apps/gcc/4.5.3/lib64 F90: gfortran Configure options: '--disable-option-checking' '--prefix=/opt/apps/gcc4_5/mvapich2/1.9b' '--with-ib-libpath=/opt/ofed/ofed/lib64/' '--with-ib-include=/opt/ofed/ofed/include/' '--with-pbs=/opt/torque/torque' '--enable-fast=O3' '--enable-shared' 'CC=gcc' 'CFLAGS= -Wl,-rpath,/opt/apps/gcc/4.5.3/lib64 -O3' 'CXX=g++' 'CXXFLAGS= -Wl,-rpath,/opt/apps/gcc/4.5.3/lib64 -O3' 'F77=gfortran' 'FFLAGS=-L/opt/ofed/ofed/lib64/ -L/lib -L/lib -Wl,-rpath,/opt/apps/gcc/4.5.3/lib64 -O3' 'FC=gfortran' '--enable-embedded-mode' '--cache-file=/dev/null' '--srcdir=.' 'LDFLAGS=-L/opt/ofed/ofed/lib64/ -L/lib -L/lib -L/lib -Wl,-rpath,/lib -L/lib -Wl,-rpath,/lib -L/lib -L/lib' 'LIBS=-libmad -lrdmacm -libumad -libverbs -ldl -lrt -lm -lpthread ' 'CPPFLAGS=-I/opt/ofed/ofed/include/ -I/opt/build/BUILD/mvapich2-1.9b/src/mpl/include -I/opt/build/BUILD/mvapich2-1.9b/src/mpl/include -I/opt/build/BUILD/mvapich2-1.9b/src/openpa/src -I/opt/build/BUILD/mvapich2-1.9b/src/openpa/src -I/opt/build/BUILD/mvapich2-1.9b/src/mpi/romio/include -I/include -I/include -I/include -I/include' Process Manager: pmi Launchers available: ssh rsh fork slurm ll lsf sge pbs manual persist Topology libraries available: hwloc Resource management kernels available: user slurm ll lsf sge pbs Checkpointing libraries available: Demux engines available: poll select > I'm particularly interested in whether mpiexec recognizes pbs as a > possible launcher or not. If it recognizes pbs as a launcher, it should > automatically select it where possible. > It is using pbs as the launcher. The problem is with the setting of the session id for the forked hydra_pmi_proxy process on mother superior. Since the hydra_pmi_proxy process has the session id set to the bash process instead mpi_exec process. As a result of this torque is unable to track the resources consumed by the process. If you look at the openmpi processes their session id is set correctly and torque is able to track the resources consumed. -- Bharath

1 0

Re: [mpich-discuss] Hydra torque support issues
by Pavan Balaji 18 Apr '13

18 Apr '13

Bharath, Can you send the output of "mpiexec -info"? I'm particularly interested in whether mpiexec recognizes pbs as a possible launcher or not. If it recognizes pbs as a launcher, it should automatically select it where possible. -- Pavan On 04/17/2013 02:21 PM US Central Time, Bharath Ramesh wrote: > On 4/17/2013 3:10 PM, Brock Palen wrote: >> Bharath, >> >> (sorry on road real quick) >> >> I do this with hydra+mvapich with Matlab and it does work. >> >> In my case I didn't find how to tell hydra when building mpich/mvapich >> to tell it where libtorque.so lives. >> >> To resolve this I download hydra alone and build it giving explicit >> options to it to find the torque library. Then you get the expected >> behavior. I don't have the directions for how I did this handy. > Thanks for the pointer, I will dig into this to see if I can get hydra > to work correctly. Currently I do pass the option > --with-pbs=/path/to/torque to the mvapich2 configure line. > > > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] Hydra torque support issues
by Bharath Ramesh 17 Apr '13

17 Apr '13

On 4/17/2013 3:10 PM, Brock Palen wrote: > Bharath, > > (sorry on road real quick) > > I do this with hydra+mvapich with Matlab and it does work. > > In my case I didn't find how to tell hydra when building mpich/mvapich to tell it where libtorque.so lives. > > To resolve this I download hydra alone and build it giving explicit options to it to find the torque library. Then you get the expected behavior. I don't have the directions for how I did this handy. Thanks for the pointer, I will dig into this to see if I can get hydra to work correctly. Currently I do pass the option --with-pbs=/path/to/torque to the mvapich2 configure line. -- Bharath

1 0