Hi Alessandro
Finally was our fault $CLASSIC_HOST must be $DPM_HOST (we were using an
old site-info configuration). Now is solved...
Cheers and thanks
Alvaro
>> Alvaro Simon Garcia ha scritto:
>>> Hi
>>>
>>> Since last day ce2.egee.cesga.es is failing sam tests because rm
>>> tests are submitting lcg-cr command with a wrong syntax:
>>>
>>> lcg-cr -v --vo ops -d -l
>>> lfn:sft-lcg-rm-cr-wn004.egee.cesga.es.071220071556.5308842
>>> file:///home/glite/opssgm001/globus-tmp.wn004.327.0/WMS_wn004_0582_https_3a_2f_2frb115.cern.ch_3a9000_2fgJRzMPXEEbK9h9scEyX19Q/work/testjob/nodes/ce2.egee.cesga.es/sft-lcg-rm-cr.txt
>>>
>>> Source file must start by 'file:' or 'gsiftp:'.
>>> lcg_cr: Invalid argument
>> Hi Alvaro,
>> the real problem is on the WN (wn004.egee.cesga.es): it doesn't see
>> the VO_OPS_DEFAULT_SE
>>
>> indeed before the test it's written
>>
>> "Move the file to the default SE () and register it with the LFN:
>> sft-lcg-rm-cr-wn004.egee.cesga.es.071220101849.5308916"
>>
>> as you can see, the SE is missing: so it seems an envirnoment
>> variable problem almost on that WN
> Hi Alessandro
>
> Thanks, that seems to be our problem, but now I wonder if we found a
> yaim configuration bug. After update to gLite 09 in prod
> VO_<VO-NAME>_DEFAULT_SE variable is missing in all users environment.
> Our site-info and vo.d directory are the same with:
>
> SW_DIR=$VO_SW_DIR/ops
> DEFAULT_SE=$CLASSIC_HOST
> STORAGE_DIR=$CLASSIC_STORAGE_DIR/ops
> ...
>
> Cheers
> Alvaro
>>
>> Cheers,
>> Alessandro
>>
>>>
>>>
>>> Instead of:
>>>
>>> lcg-cr -v --vo ops -d *se2.egee.cesga.es* -l
>>> lfn:sft-lcg-rm-cr-wn001.egee.cesga.es.071219022149.2130797
>>> file:///home/glite/opssgm001/globus-tmp.wn001.31420.0/WMS_wn001_031665_https_3a_2f_2frb104.cern.ch_3a9000_2f3z7xkxCjEVdVNLSRAzYqWA/work/testjob/nodes/ce2.egee.cesga.es/sft-lcg-rm-cr.txt
>>>
>>>
>>>
>>> There are other sites with the same problem?
>>>
>>> Cheers and thanks
>>> Alvaro
>>
>>
>
>
|