On May 5, 2005, at 12:32 AM, gsaxena_at_[hidden] wrote:
> Question for LAM members:
> Is there a signalling mechanism through which the
> root process may be able to notify other nodes to quit their
> jobs and finalize ?
>
I think David Cronk had a nice response to this question in your
other e-mail thread. You're best off using "control" messages
between MPI applications
> An interesting far-sight:
> any thoughts as to how error handling be addressed in such an
> application?
>
Take a look at the LAM example "fault" in the LAM source tarball
(lam-7.1.x/examples/fault/). It should do about what you are looking
for.
Hope this helps,
Brian
|