On Fri, 22 Aug 2003, wei zhang wrote:
> (1) When I did lamboot my_host_file, I got the following error message.
> ----------------------------------------------------------------------------
> It seems that LAM was not able to remove a directory properly. This should
> not happen, and will probably require manual intervention on your part.
> tkill was trying to remove the following directory:
>
> ./lam-cfd_at_[hidden]
> [snipped]
> ----------------------------------------------------------------------------
This is an odd error, and suggests a higher-level problem. I'm concerned
that it was unable to remove the directory, and I'm concerned that your
session directory was created off ".".
A few questions:
- Do you have $TMPDIR set to "."?
- Is "." a network-mounted filesystem?
- Is `pwd` of "." available on all nodes that you're trying to run on?
This is all with the big disclaimer assuming that you can't just remove
this directory manually and run lamboot successfully. If you *can*
manually remove the directory and run lamboot successfully, then this is
all moot. :-)
> (4) When I ran " mpirun C hello", I got following massage.
> ----------------------------------------------------------------------------
> It seems that there is no lamd running on the host master.xx.xxxxx.com.
> [snipped]
This is probably likely, given that lamboot failed.
--
{+} Jeff Squyres
{+} jsquyres_at_[hidden]
{+} http://www.lam-mpi.org/
|
