Hello peeps,
Our mighty SRM, fal-pygrid-20.lancs.ac.uk, keeps stopping publishing
its space dynamically at regular intervals for some strange reason. We
timed these intervals and found that they came roughly every 8 hours
and stopped when the edg-mkgridmap cron job ran. Using this we found
we could manually restart the dynamic space publishing by invoking a:
/opt/d-cache/bin/grid-mapfile2dcache-kpwd
However artificially creating a cronjob that ran this command
imbetween each mkgridmap failed to fix it.
I think I remember something similar being mentioned way back in the
mists of time on this list but I could be mistaken.
Any ideas what's going wrong here? We could just get edg-mkgridmap to
run more often but it would be nice to know what's going on.
cheers,
matt
|