[mpich-discuss] mpdboot error message

[Dave Goodell]

I suspect you are hitting this bug: https://trac.mcs.anl.gov/projects/mpich2/ticket/974

"mpdboot" is normally issued as a foreground command, not in the  
background.  It should exit immediately after setting up the ring.  If  
your mpdboot is hanging then that is the real bug.  My access to some  
CentOS machines to reproduce/fix always seems to be a few days away,  
so I haven't been able to fix it yet.

In the meantime, my recommendation is to use hydra unless you need one  
of the very few mpd-specific features that still exist:

http://wiki.mcs.anl.gov/mpich2/index.php/Using_the_Hydra_Process_Manager

Out of curiosity, what operating system are you using?  CentOS/RHEL?

-Dave

On Jan 15, 2010, at 11:02 AM, Cezary Śliwa wrote:

>
> Hello,
>
> mpdboot in mpich2-1.2.1 displays an error message when mpdallexit is  
> used:
>
> sl2klast2-n5 ~$ mpdboot -v -n 2 &
> [1] 6793
> sl2klast2-n5 ~$ running mpdallexit on sl2klast2-n5
> LAUNCHED mpd on sl2klast2-n5  via
> RUNNING: mpd on sl2klast2-n5
> LAUNCHED mpd on sl2klast2-n6  via  sl2klast2-n5
>
> sl2klast2-n5 ~$ mpdallexit
> sl2klast2-n5 ~$ mpdboot_sl2klast2-n5 (handle_mpd_output 415): failed  
> to connect to mpd on sl2klast2-n6
>
> [1]+  Exit 255                mpdboot -v -n 2
> sl2klast2-n5 ~$ cat mpd.hosts
> sl2klast2-n5:4
> sl2klast2-n6:4
>
>
> Cezary Sliwa
>
>
> _______________________________________________
> mpich-discuss mailing list
> mpich-discuss at mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss