[mpich-discuss] good system gone bad..

[Darius Buntinas]

Check that the hydra proxy installed (and accessible) on every node.

-d

On Dec 6, 2010, at 4:27 PM, SULLIVAN David (AREVA) wrote:

> All,
>  
> I am trying to troubleshoot a problem with my MPICH2 install. I compiled version 1.30 with gcc and intel fortran. When I try to execute anything I am greeted with the following:
>  
> bash: /global/mpich2-1.3/bin/hydra_pmi_proxy: No such file or directory
> It is lying though:
> 
> total 2916
> -rwxr-xr-x 1 root root   1656 Dec  6 16:54 bt2line
> -rwxr-xr-x 1 root root  10008 Dec  6 16:54 check_callstack
> -rwxr-xr-x 1 root root  56622 Dec  6 16:54 clog2_join
> -rwxr-xr-x 1 root root   1970 Dec  6 16:54 clog2print
> -rwxr-xr-x 1 root root  53798 Dec  6 16:54 clog2_print
> -rwxr-xr-x 1 root root  53799 Dec  6 16:54 clog2_repair
> -rwxr-xr-x 1 root root   1959 Dec  6 16:54 clog2TOslog2
> -rwxr-xr-x 1 root root   1960 Dec  6 16:54 clogprint
> -rwxr-xr-x 1 root root   1956 Dec  6 16:54 clogTOslog2
> -rwxr-xr-x 1 root root  84887 Dec  6 16:54 hwloc-bind
> -rwxr-xr-x 1 root root  83889 Dec  6 16:54 hwloc-calc
> -rwxr-xr-x 1 root root  79196 Dec  6 16:54 hwloc-distrib
> lrwxrwxrwx 1 root root      6 Dec  6 16:54 hwloc-info -> lstopo
> lrwxrwxrwx 1 root root      6 Dec  6 16:54 hwloc-ls -> lstopo
> lrwxrwxrwx 1 root root     10 Dec  6 16:54 hwloc-mask -> hwloc-calc
> -rwxr-xr-x 1 root root 422199 Dec  6 16:54 hydra_nameserver
> -rwxr-xr-x 1 root root 420869 Dec  6 16:54 hydra_persist
> -rwxr-xr-x 1 root root 586873 Dec  6 16:54 hydra_pmi_proxy
> -rwxr-xr-x 1 root root   1946 Dec  6 16:54 jumpshot
> -rwxr-xr-x 1 root root   1944 Dec  6 16:54 logconvertor
> -rwxr-xr-x 1 root root 136310 Dec  6 16:54 lstopo
> lrwxrwxrwx 1 root root      6 Dec  6 16:54 mpic++ -> mpicxx
> -rwxr-xr-x 1 root root   8974 Dec  6 16:54 mpicc
> -rwxr-xr-x 1 root root   8874 Dec  6 16:54 mpich2version
> -rwxr-xr-x 1 root root   8659 Dec  6 16:54 mpicxx
> lrwxrwxrwx 1 root root     13 Dec  6 16:54 mpiexec -> mpiexec.hydra
> -rwxr-xr-x 1 root root 748519 Dec  6 16:54 mpiexec.hydra
> -rwxr-xr-x 1 root root  10640 Dec  6 16:54 mpif77
> -rwxr-xr-x 1 root root  12615 Dec  6 16:54 mpif90
> lrwxrwxrwx 1 root root     13 Dec  6 16:54 mpirun -> mpiexec.hydra
> -rwxr-xr-x 1 root root   3430 Dec  6 16:54 parkill
> -rwxr-xr-x 1 root root  20427 Dec  6 16:54 plpa-info
> -rwxr-xr-x 1 root root  40775 Dec  6 16:54 plpa-taskset
> -rwxr-xr-x 1 root root   1965 Dec  6 16:54 slog2filter
> -rwxr-xr-x 1 root root   1983 Dec  6 16:54 slog2navigator
> -rwxr-xr-x 1 root root   2015 Dec  6 16:54 slog2print
> -rwxr-xr-x 1 root root   1974 Dec  6 16:54 slog2updater
> clearly it IS there, and it is accessible :
>  
> [dsullivan at athos bin]$ which hydra_pmi_proxy
> /global/mpich2-1.3/bin/hydra_pmi_proxy
> [dsullivan at athos bin]$
> This all worked this morning and little has changed since (a restart maybe). Any direction would be greatly appreciated.
>  
>  
> Regards,
>  
> David Sullivan
>  
>  
>  
> AREVA NP INC
> 400 Donald Lynch Boulevard
> Marlborough, MA, 01752
> Phone: (508) 573-6721
> Fax: (434) 382-5597
> David.Sullivan at AREVA.com
>  
> The information in this e-mail is AREVA property and is intended solely for the addressees. Reproduction and distribution are prohibited. Thank you .
>  
> _______________________________________________
> mpich-discuss mailing list
> mpich-discuss at mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss