Discussion related to cado-nfs

[Paul Leyland <paul.leyland@gmail.com>]

That's good.

My plan was to write a Perl function along these lines:

sub sanity_clause($$$$)
{
    my ($host, $bindir, $tmpdir, $wdir) = @_;
 
    my @binaries = ('sieve/las',
                    'sqrt/sqrt',
                    # Etc
                   );

   if ($host eq 'localhost') {
        -d $bindir && -r $bindir && -x $bindir or die "$bindir is 
inaccesible\n"
        foreach (@binaries) {
            -x "$bindir/$_ or die "$bindir/$_ is not executable\n";
        }

# Test $tmpdir and "wdir for read/writable directories or die.

    } else {

# Use ssh to perform analogous tests on the remote hosts
# but warn rather than die.

#Set a flag in an array @accessible_hosts if all tests pass
# and then use that to decide whether to run jobs there.

    }
}

and , of course, to call sanity_clause() for each of the machines which
have at least one core set in mach_desc.  Even after all that has been
done, explicit permission should be needed for deletion of relations to
take place.

Paul

On Sun, 2013-04-07 at 22:20 +0200, Zimmermann Paul wrote:
> Paul,
> 
> > I tried to add another machine as a siever but didn't really understand
> > the format of the mach_desc file and, in particular, the scope of the
> > bindir declaration.  When cadofactor.pl restarted it tried to de-dup and
> > purge the relations found so far but used the wrong path.  Instead of
> > dying with an error "program not found" or similar, it deleted all the
> > relation files.
> 
> we are aware of that bug:
> 
> https://gforge.inria.fr/tracker/index.php?func=detail&aid=13209&group_id=2065&atid=7442
> 
> It should be fixed in the new version of the cadofactor script that Alex is
> currently designing.
> 
> The idea is that relations should not be deleted without an explicit
> "YES" from the user.
> 
> Paul