Although I had to nic a grid-mapfile from Steve (mkgridmap-cron doesn't do
anything for me at the moment), I have just managed the following:
[stokes@pptb01 stokes]$ globus-job-run tbce01 /bin/hostname
tbce01.physics.ox.ac.uk
which, while I'm sure not exciting to most, almost left me in tears of
joy. The problem is that the same cannot be repeated from gppui04. It
works fine to other sites (for instance, QMUL), but just sits there when I
try it for tbce01 (10 minutes has been the most time I've given it so
far).
Any suggestions for what might cause this would be greatly appreciated.
To get this working, I had to comment out lcas_userban.mod and
lcas_timeslots.mod in /opt/edg/etc/lcas/lcas.db. This was inspired by the
failure messages which were coming up in /var/logs/globus-gatekeeper.log:
Notice: 0: lcasmod_name = /opt/edg/lib/lcas/lcas.mod
LCAS 5: Initialization LCAS version 1.1.8
LCAS 0: lcas.mod-lcas_init(): Reading LCAS database
/opt/edg/etc/lcas/lcas.db
LCAS 0: lcas_timeslots.mod-plugin_initialize() error: timeslots
file required !
LCAS 0: lcas.mod-PluginInit(): "plugin_initialize()" failed
LCAS 0: lcas.mod-lcas_init(): error initializing plugin:
lcas_timeslots.mod
Failure: LCAS initialization failure.
I don't know what other logs to look in (either on gppui04 or on tbce01)
to see why the globus-job-run isn't returning anything, and there doesn't
seem to be a "verbose" or "debug" mode either -- any help on this would be
appreciated.
Cheers,
Ian.
--
Ian Stokes-Rees [log in to unmask]
Particle Physics, Oxford http://www-pnp.physics.ox.ac.uk/~stokes
|