The executables must be available on all nodes - normally, this is done by
putting them in an NFS-mounted directory.
On Dec 29, 2009, at 6:35 AM, vishal shorghar wrote:
> HI All,
>
> Today I reran the trivial test on two nodes with (via the --scratch option)
> to a NFS share that is accessible to all nodes in hostlist (as suggested by
> Ethan).But still no luck.
> I have shared
> "/root/mtt-svn/samples/installs/nRpF/tests/trivial/test_get__trivial" on my
> head node which is accessible to client node.
> Then I ran test as "cat samples/developer.ini samples/trivial.ini |
> client/mtt --scratch
> /root/mtt-svn/samples/installs/nRpF/tests/trivial/test_get__trivial --verbose
> - " but still its giving me the same error which I was getting earlier.
>
> We copied /root/mtt-svn/samples/installs/nRpF/tests/trivial/test_get__trivial
> directory on the client node and changed the trivial.ini as follows:
> -- simple_only:tests = &find_executables(".")
> ++ simple_only:tests =
> &find_executables("/root/mtt-svn/samples/installs/nRpF/tests/trivial/test_get__trivial")
>
> and after this we ran the test and we were able to run it successfully.
>
> Does this mean we manually have to either copy the test executables to client
> nodes or we have to mount the test executables directory?
>
> Can you please guide us on this to proceed further.
>
> Regards,
> Vishal
>
>
>
>
>
>
> > Hi All,
> > >
> > > I have one issue with MTT trivial tests.All tests are not getting
> > > passed,Please read below for detailed description.
> > >
> > > Today I ran mtt trivial tests with latest ofed package
> > > OFED-1.5-20091217-0600 (ompi-1.4), between two machines,I was able to
> > > run
> > > the MTT trivial tests manually but not through MTT framework. I think
> > > we
> > > are missing some configuration steps since it is unable to find the
> > > test
> > > executables in the test run phase of the MTT.
> > >
> > > -> When we ran it through MTT it gave us the error and exits.
> > > I ran the test as "cat developer.ini trivial.ini | ../client/mtt
> > > --verbose - "
> > >
> > > -> When we analyzed error from
> > > /root/mtt-svn/samples/Test_Run-trivial-my_installation-1.4.txt file we
> > > found it is not getting the executable files of the different test to
> > > execute.
> > >
> > > -> Then we found that those executables were being generated only on
> > > one
> > > of the machine out of two machines. So, we manually copied the tests
> > > from
> > >
> > > /root/mtt-svn/samples/installs/nRpF/tests/trivial/test_get__trivial/c_ring
> > > to another machine.
> > >
> > > -> And we ran it manually as shown below and it worked fine:
> > > mpirun --host 102.77.77.64,102.77.77.68 -np 2 --mca btl openib,sm,self
> > > --prefix
> > >
> > > /usr/mpi/gcc/openmpi-1.4/root/mtt-svn/samples/installs/nRpF/tests/trivial/test_get__trivial/c_ring
> > >
> > > -> I am attaching file trivial.ini,developer.ini and
> > > /root/mtt-svn/samples/Test_Run-trivial-my_installation-1.4.txt.
> > >
> > > Let us know if I am missing some configuration steps.
> > >
>
> You need to set your scratch directory (via the --scratch option) to
> an NFS share that is accessible to all nodes in your hostlist. MTT
> won't copy local files onto each node for you.
>
> Regards,
> Ethan
>
> #
> # Copyright (c) 2007 Sun Microystems, Inc. All rights reserved.
> #
>
> # Template MTT configuration file for Open MPI developers. The intent
> # for this template file is to establish at least some loose
> # guidelines for what Open MPI core developers should be running
> # before committing changes to the OMPI repository. This file is not
> # intended to be an exhaustive sample of all possible fields and
> # values that MTT offers. Each developer will undoubtedly have to
> # edit this template for their own needs (e.g., pick compilers to use,
> # etc.), but this file provides a baseline set of configurations that
> # we intend for you to run.
> #
> # Sample usage:
> # cat developer.ini intel.ini | client/mtt -
> alreadyinstalled_dir=/your/install
> # cat developer.ini trivial.ini | client/mtt -
> alreadyinstalled_dir=/your/install
> #
>
> [MTT]
> # No overrides to defaults
>
> # Fill this field in
>
> #hostlist = 102.77.77.63 102.77.77.54 102.77.77.64 102.77.77.68
> #hostlist = 102.77.77.66 102.77.77.68 102.77.77.63 102.77.77.64 102.77.77.53
> 102.77.77.54 102.77.77.243 102.77.77.65
> hostlist = 102.77.77.64 102.77.77.63
> hostlist_max_np = 2
> max_np = 2
> force = 1
> #prefix = /usr/mpi/gcc/openmpi-1.4/bin
>
> #----------------------------------------------------------------------
>
> [MPI Details: Open MPI]
>
> exec = mpirun @hosts@ -np &test_np() @mca@ --prefix &test_prefix()
> &test_executable() &test_argv()
>
> mca = --mca btl openib,sm,self
>
> hosts = <<EOT
> &if(
> &have_hostfile(),
> &stringify("--hostfile ", &hostfile()),
> &if(
> &have_hostlist(),
> &stringify("--host ", &hostlist()),
> ""
> )
> )
> EOT
>
> # Here is a good place to put any cleanup commands (e.g.,
> # kill a stale process, remove a leftover file, etc.)
> after_all_exec = &shell("echo done.")
>
> #----------------------------------------------------------------------
>
> [MPI get: My Installation]
> mpi_details = Open MPI
> # Leave this parameter blank to
> # have MTT search your path for an MPI
> # alreadyinstalled_dir = /your/installation
> module = AlreadyInstalled
>
> #----------------------------------------------------------------------
>
> [MPI install: My Installation]
> mpi_get = My Installation
> module = Analyze::OMPI
>
> #----------------------------------------------------------------------
>
> [Reporter: text file backup]
> module = TextFile
>
> textfile_filename = $phase-$section-$mpi_name-$mpi_version.txt
>
> # User-defined report headers/footers
> textfile_summary_header = <<EOT
> hostname: &shell("hostname")
> uname: &shell("uname -a")
> who am i: &shell("who am i")
> EOT
>
> textfile_summary_footer =
> textfile_detail_header =
> textfile_detail_footer =
>
> textfile_textwrap = 78
>
>
> +--------------------------+------------------------------------------------------------------------------+
> | Field | Value
> |
> +--------------------------+------------------------------------------------------------------------------+
> | description |
> |
> | environment |
> |
> | exit_signal | -1
> |
> | exit_value | 131
> |
> | launcher | mpirun
> |
> | mpi_install_section_name | my installation
> |
> | mpi_name | my installation
> |
> | mpi_version | 1.4
> |
> | network |
> |
> | np | 2
> |
> | parameters |
> |
> | phase | Test Run
> |
> | resource_manager | none
> |
> | result_message | Failed; exit status: 131
> |
> | result_stderr |
> |
> | test_build_section_name | trivial
> |
> | variant | 1
> |
> | |
> |
> | command | mpirun --host 102.77.77.64,102.77.77.63 -np
> 2 --mca btl openib,sm,self |
> | | --prefix /usr/mpi/gcc/openmpi-1.4/ ./f77_hello
> |
> | duration | 1 seconds
> |
> | result_stdout |
> -------------------------------------------------------------------------- |
> | | mpirun was unable to launch the specified
> application as it could not access |
> | | or execute an executable:
> |
> | |
> |
> | | Executable: ./f77_hello
> |
> | | Node: 102.77.77.63
> |
> | |
> |
> | | while attempting to start process rank 1.
> |
> | |
> -------------------------------------------------------------------------- |
> | start_timestamp | 1262064675
> |
> | start_timestamp_human | Tue Dec 29 05:31:15 2009
> |
> | test_name | f77_hello
> |
> | |
> |
> | command | mpirun --host 102.77.77.64,102.77.77.63 -np
> 2 --mca btl openib,sm,self |
> | | --prefix /usr/mpi/gcc/openmpi-1.4/ ./f90_ring
> |
> | duration | 2 seconds
> |
> | result_stdout |
> -------------------------------------------------------------------------- |
> | | mpirun was unable to launch the specified
> application as it could not access |
> | | or execute an executable:
> |
> | |
> |
> | | Executable: ./f90_ring
> |
> | | Node: 102.77.77.63
> |
> | |
> |
> | | while attempting to start process rank 1.
> |
> | |
> -------------------------------------------------------------------------- |
> | start_timestamp | 1262064676
> |
> | start_timestamp_human | Tue Dec 29 05:31:16 2009
> |
> | test_name | f90_ring
> |
> | |
> |
> | command | mpirun --host 102.77.77.64,102.77.77.63 -np
> 2 --mca btl openib,sm,self |
> | | --prefix /usr/mpi/gcc/openmpi-1.4/ ./f90_hello
> |
> | duration | 1 seconds
> |
> | result_stdout |
> -------------------------------------------------------------------------- |
> | | mpirun was unable to launch the specified
> application as it could not access |
> | | or execute an executable:
> |
> | |
> |
> | | Executable: ./f90_hello
> |
> | | Node: 102.77.77.63
> |
> | |
> |
> | | while attempting to start process rank 1.
> |
> | |
> -------------------------------------------------------------------------- |
> | start_timestamp | 1262064678
> |
> | start_timestamp_human | Tue Dec 29 05:31:18 2009
> |
> | test_name | f90_hello
> |
> | |
> |
> | command | mpirun --host 102.77.77.64,102.77.77.63 -np
> 2 --mca btl openib,sm,self |
> | | --prefix /usr/mpi/gcc/openmpi-1.4/ ./f77_ring
> |
> | duration | 1 seconds
> |
> | result_stdout |
> -------------------------------------------------------------------------- |
> | | mpirun was unable to launch the specified
> application as it could not access |
> | | or execute an executable:
> |
> | |
> |
> | | Executable: ./f77_ring
> |
> | | Node: 102.77.77.63
> |
> | |
> |
> | | while attempting to start process rank 1.
> |
> | |
> -------------------------------------------------------------------------- |
> | start_timestamp | 1262064679
> |
> | start_timestamp_human | Tue Dec 29 05:31:19 2009
> |
> | test_name | f77_ring
> |
> | |
> |
> | command | mpirun --host 102.77.77.64,102.77.77.63 -np
> 2 --mca btl openib,sm,self |
> | | --prefix /usr/mpi/gcc/openmpi-1.4/ ./cxx_ring
> |
> | duration | 1 seconds
> |
> | result_stdout |
> -------------------------------------------------------------------------- |
> | | mpirun was unable to launch the specified
> application as it could not access |
> | | or execute an executable:
> |
> | |
> |
> | | Executable: ./cxx_ring
> |
> | | Node: 102.77.77.63
> |
> | |
> |
> | | while attempting to start process rank 1.
> |
> | |
> -------------------------------------------------------------------------- |
> | start_timestamp | 1262064680
> |
> | start_timestamp_human | Tue Dec 29 05:31:20 2009
> |
> | test_name | cxx_ring
> |
> | |
> |
> | command | mpirun --host 102.77.77.64,102.77.77.63 -np
> 2 --mca btl openib,sm,self |
> | | --prefix /usr/mpi/gcc/openmpi-1.4/ ./cxx_hello
> |
> | duration | 2 seconds
> |
> | result_stdout |
> -------------------------------------------------------------------------- |
> | | mpirun was unable to launch the specified
> application as it could not access |
> | | or execute an executable:
> |
> | |
> |
> | | Executable: ./cxx_hello
> |
> | | Node: 102.77.77.63
> |
> | |
> |
> | | while attempting to start process rank 1.
> |
> | |
> -------------------------------------------------------------------------- |
> | start_timestamp | 1262064681
> |
> | start_timestamp_human | Tue Dec 29 05:31:21 2009
> |
> | test_name | cxx_hello
> |
> | |
> |
> | command | mpirun --host 102.77.77.64,102.77.77.63 -np
> 2 --mca btl openib,sm,self |
> | | --prefix /usr/mpi/gcc/openmpi-1.4/ ./c_hello
> |
> | duration | 1 seconds
> |
> | result_stdout |
> -------------------------------------------------------------------------- |
> | | mpirun was unable to launch the specified
> application as it could not access |
> | | or execute an executable:
> |
> | |
> |
> | | Executable: ./c_hello
> |
> | | Node: 102.77.77.63
> |
> | |
> |
> | | while attempting to start process rank 1.
> |
> | |
> -------------------------------------------------------------------------- |
> | start_timestamp | 1262064683
> |
> | start_timestamp_human | Tue Dec 29 05:31:23 2009
> |
> | test_name | c_hello
> |
> | |
> |
> | command | mpirun --host 102.77.77.64,102.77.77.63 -np
> 2 --mca btl openib,sm,self |
> | | --prefix /usr/mpi/gcc/openmpi-1.4/ ./c_ring
> |
> | duration | 1 seconds
> |
> | result_stdout |
> -------------------------------------------------------------------------- |
> | | mpirun was unable to launch the specified
> application as it could not access |
> | | or execute an executable:
> |
> | |
> |
> | | Executable: ./c_ring
> |
> | | Node: 102.77.77.63
> |
> | |
> |
> | | while attempting to start process rank 1.
> |
> | |
> -------------------------------------------------------------------------- |
> | start_timestamp | 1262064684
> |
> | start_timestamp_human | Tue Dec 29 05:31:24 2009
> |
> | test_name | c_ring
> |
> | |
> |
> +--------------------------+------------------------------------------------------------------------------+
>
> #
> # Copyright (c) 2007 Sun Microystems, Inc. All rights reserved.
> #
>
> #
> # This INI file is an addendum to samples/developer.ini (which only
> # contains Trivial tests).
> #
> # Sample usage:
> # cat developer.ini trivial.ini | client/mtt -
> alreadyinstalled_dir=/your/install
> #
>
> [Test get: trivial]
> module = Trivial
>
> #----------------------------------------------------------------------
>
> [Test build: trivial]
> test_get = trivial
> save_stdout_on_success = 1
> merge_stdout_stderr = 1
> stderr_save_lines = 100
>
> module = Trivial
>
> #----------------------------------------------------------------------
>
> [Test run: trivial]
> test_build = trivial
> pass = &and(&cmd_wifexited(), &eq(&cmd_wexitstatus(), 0))
> timeout = &test_np()
> save_stdout_on_pass = 1
> merge_stdout_stderr = 1
> stdout_save_lines = 100
> np = &env_max_procs()
>
> specify_module = Simple
> simple_only:tests = &find_executables(".")
>
> #----------------------------------------------------------------------
>
> _______________________________________________
> users mailing list
> us...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/users
