Hi,
I have been trying to run some CMS MC on the grid. The jobs run all over
the place and then copy the ouput back to the SE at RAL ... the same one
that I was trying to delete from earlier... infact these files have the
same names as the ones that were deleted earlier.
Of the 24 jobs that ran the MC correctly. Only 4 managed to copy their
output to the SE. All of these also successfully registered the output
files with the RLS.
All of the others failed.
From the LogFile the command were of the form:
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
> lcg-cr --vo cms
file:/grid/fzk.de/mounts/nfs/home/cms001/globus-tmp.c01-012-122.23544.0/WMS_c01-012-122_024123_https_3a_2f_2flcgrb01.gridpp.rl.ac.
uk_3a9000_2fQgjsHu8YjBtSfJCxb3qqDA/sm05_wbb_lv_toprex_215800017.ntpl -d
srm://dcache.gridpp.rl.ac.uk:8443/pnfs/gridpp.rl.ac.uk/data/cms/cms/ProdLCG/
CMKIN_4_4_0/sm05_wbb_lv_toprex/sm05_wbb_lv_toprex_215800017.ntpl -l
lfn:sm05_wbb_lv_toprex_215800017.ntpl
StageOutFiles: failed to copy and register
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
The command was tried 3 times in each script and failed on all 3 attempts.
I have spent the vast majority of my time in GridPP c9oncerned with
workload and this is my first attempt in some long while to use the data
moving tools.
So my questions are:
1. Is this typical?
2. Is the problem with the RLS or the SE?
3. Is there anyway to tell?
4. Is there anything that I can do about
5. Could this problem be caused by the them sharing the same names as the
ones I deleted earlier?
Ok... I will re-run these jobs...
ttfn,
david
|