On Fri, Sep 09, 2005 at 09:42:23AM +0100 or thereabouts, Burke, S (Stephen) wrote:
> LHC Computer Grid - Rollout
> > [mailto:[log in to unmask]] On Behalf Of Steve Traylen
> said:
> > A few people have noticed this, the RAL one is published about 4 times
> > in the UK for instance. I did consider asking people not to publish it
> > but if the RAL site BDII goes down our RB endpoint is still
> > published so this is good thing.
>
> I don't really see that, if the RAL BDII is down the RB may well be down
> too. And what happens if the information changes, how do you make sure
> you propagate changes to random sites which are duplicating it? Some
> kind of managed backup might make sense (e.g. R-GMA generally has
> several archivers) but I don't think it's a good idea to get it from
> random misconfiguration ...
Each of the site BDIIs are polling the RBs and other GRISes. If the RAL RB
goes down it will vanish from RAL's site BDII and Manchester's BDII.
Steve
>
> Stephen
--
Steve Traylen
[log in to unmask]
http://www.gridpp.ac.uk/
|