On Mon, 28 Mar 2005, Kyriacos Neocleous wrote:
> Early this month we faced a hardware problem on se101.grid.ucy.ac.cy:
> two of the four RAID 5 disks went offline simultaneously due to raid
> controller failure, and we were unable to recover the data, no matter
> what we tried (in collaboration with the Promise support team, raid
> controller manufacturer). The data was mostly dteam replication test
> files, but a few belonged to the lhcb VO, while I believe there were no
> replicas from other VOs when the system failed.
Just in case ...
You can easily check for the replicas which were located on your dead
system, using edg-lrc command. For example, if you'd like to know which
files belonging to LHCb VO were located on your SE, you can issue the
following command:
edg-lrc mappingsByPfn "sfn://se101.grid.ucy.ac.cy/*" --endpoint http://rlslhcb.cern.ch:7777/lhcb/v2.2/edg-local-replica-catalog/services/edg-local-replica-catalog
and similarly for other VOs (replacing the endpoint with URL to correct
RLS/RMC catalogue for given VO).
--
Jiri Kosina
Institute of Physics, Academy of sciences of the Czech Republic
|