[mpich-discuss] mpiexec (hydra) can't run in the background

[Manhui Wang]

Hello,

I built the mpich2-1.3.2 library with ifort/icc/icpc 11.1.072 using the
following options:

nice -n +18 ./configure  --with-device=ch3:nemesis
--prefix=/path/mpich2-1.3.2-nemesis-install FC=ifort --enable-fc
F77=ifort --enable-f77 CC=icc CXX=icpc --enable-cxx 2>&1 | tee configure.log

nice -n +18 make 2>&1 | tee make.log

nice -n +18 make install 2>&1 | tee install.log

Reproduce the problem with:

mpicc hello_world_wait.c -o hello.1.3.2
mpiexec -np 2 ./hello.1.3.2 > output132.log&

then immediately try any command eg.

echo hello

then it fails with error message like

 [mpiexec at arccacluster8] HYDU_sock_read (./utils/sock/sock.c:222): read
errno (Input/output error)
[mpiexec at arccacluster8] econtrol_cb (./pm/pmiserv/pmiserv_cb.c:249):
assert (!closed) failed
[mpiexec at arccacluster8] HYDT_dmxu_poll_wait_for_event
(./tools/demux/demux_poll.c:77): callback returned error status
[mpiexec at arccacluster8] HYD_pmci_wait_for_completion
(./pm/pmiserv/pmiserv_pmci.c:206): error waiting for event
[mpiexec at arccacluster8] main (./ui/mpich/mpiexec.c:404): process manager
error waiting for completion
cho hello
hello
[1]+  Exit 255                mpiexec -np 2 ./hello.1.3.2 > output132.log


Thanks,
Manhui


Pavan Balaji wrote:
> Hello,
> 
> On 02/11/2011 03:13 AM, Manhui Wang wrote:
>>   mpicc hello_world_wait.c -o hello.1.3.2
>> [sacmw4 at arccacluster11 testprogram]$ mpiexec -np 2 ./hello.1.3.2
>>> output132.log&
> 
> I just tried the 1.3.2 release on Ubuntu Maverick with gcc 4.4.3 and it
> seems to work fine. It's hard to tell what's going wrong with the
> information provided.
> 
> Would you be willing to try out other snapshots from
> http://www.mcs.anl.gov/research/projects/mpich2/downloads/tarballs/nightly/hydra
> and tell me approximately which version this broke at?
> 
> Thanks,
> 
>  -- Pavan
> 

-- 
-----------
Manhui  Wang
School of Chemistry, Cardiff University,
Main Building, Park Place,
Cardiff CF10 3AT, UK