Hi Daniela,
Thanks a lot for testing this!
Right now there is no large job submission expected for LSST, especially
not before the end of the year.
LSST will start with some work on GridPP early next year when we try to
port another workflow to Grid jobs. This will most likely also involve
copying new input data files into GridPP.
It could be that there are some test jobs in the next days/weeks to get
familiar with the grid job submission, but that will be based on the
previous work flow Joe used and reuse the current data files.
Thanks,
Marcus
On Thu, 8 Dec 2016, Daniela Bauer wrote:
> Hi Marcus et al,
>
> we tested xrootd access on the test server.
> It worked, but it required a number of quite substantial updates.
> I'm somewhat reluctant to do this on the production server before
> Christmas, especially as we are currently debugging dune. As all the LSST
> files are at Imperial and LSST is not doing a production run, I hope this
> works.
> If we get around to it any earlier of course we'll enable it earlier.
>
> Regards,
> Daniela
>
> On 8 December 2016 at 10:34, Marcus Ebert <[log in to unmask]> wrote:
>
>> Hi Daniela,
>>
>> Yes, I think I removed all the files on gridpp09.ecdf.ed.ac.uk and
>> disabled LSST support locally. It can be removed from the configuration
>> then.
>> Only srm.glite.ecdf.ed.ac.uk (UKI-SCOTGRID-ECDF-disk) should remain in
>> the system.
>>
>> Cheers,
>> Marcus
>>
>>
>> On Thu, 8 Dec 2016, Daniela Bauer wrote:
>>
>> Hi,
>>>
>>> So we are all done with this SE and I can remove it from the dirac
>>> configuration ?
>>> Regards,
>>> Daniela
>>>
>>> On 7 December 2016 at 15:00, Marcus Ebert <[log in to unmask]> wrote:
>>>
>>> Hi Alessandra,
>>>>
>>>> That worked perfect using dirac-dms-remove-catalog-replicas to remove
>>>> the
>>>> entries!
>>>>
>>>> Thanks,
>>>> Marcus
>>>>
>>>>
>>>> On Wed, 7 Dec 2016, Alessandra Forti wrote:
>>>>
>>>> Hi,
>>>>
>>>>>
>>>>> you should be able to delete the replicas and the entries in the
>>>>> catalogue (dirac-dms-remove-replicas/files) at the same time. If you
>>>>> don't want to do that because it will take significantly longer you
>>>>> should
>>>>> be able to remove only the entries in the catalogue
>>>>> (dirac-dms-remove-catalog-replias/files).
>>>>>
>>>>> cheers
>>>>> alessandra
>>>>>
>>>>> On 07/12/2016 14:02, Marcus Ebert wrote:
>>>>>
>>>>> Hi Daniela,
>>>>>>
>>>>>> Ah, right. Removing the files from the catalogue still needs to be
>>>>>> done.
>>>>>> I could just rename the directory to lsst and should be available and
>>>>>> possible to delete the files.
>>>>>> However, I would like to see first if it's possible to just remove the
>>>>>> entries of the replicas in the file catalogue. This could be helpful
>>>>>> if
>>>>>> a
>>>>>> server on a site would fail and files would no longer be available (to
>>>>>> be
>>>>>> deleted).
>>>>>>
>>>>>> There is the dirac-dms-set-replica-status tool available. Do you know
>>>>>> if
>>>>>> that could be used to set the status for the replicas on a specific
>>>>>> site
>>>>>> to "deleted"? Unfortunately, the help option of this command doesn't
>>>>>> show
>>>>>> what values can be specified for "status".
>>>>>> Or is there another command that could be used to remove the entries
>>>>>> of
>>>>>> specific replicas/sites from the file catalogue?
>>>>>>
>>>>>>
>>>>>> Cheers,
>>>>>> Marcus
>>>>>>
>>>>>> On Wed, 7 Dec 2016, Daniela Bauer wrote:
>>>>>>
>>>>>> Hi Marcus,
>>>>>>>
>>>>>>>> as far as I can tell the lsst files on gridpp09.ecdf.ed.ac.uk are
>>>>>>>>
>>>>>>> still
>>>>>>
>>>>>>> registered in the dirac file catalogue.
>>>>>>> If they are still on the SE and you are able to put the directoty
>>>>>>>
>>>>>> back, > you
>>>>>>
>>>>>>> can delete and unregister them at the same time using this script:
>>>>>>> https://raw.githubusercontent.com/ic-hep/DIRAC-tools/master/
>>>>>>>
>>>>>> recursiverm.py > > > Regards,
>>>>>>
>>>>>>> Daniela
>>>>>>>
>>>>>>>> On 7 December 2016 at 10:52, Marcus Ebert <[log in to unmask]>
>>>>>>>>
>>>>>>> wrote:
>>>>>>
>>>>>>> Hi,
>>>>>>>>>
>>>>>>>>>> Just an update on this:
>>>>>>>>>> On Thu, 1 Dec 2016, Daniela Bauer wrote:
>>>>>>>>>> Hi Marcus,
>>>>>>>>>>
>>>>>>>>>>> I've replicated the content from UKI-SCOTGRID-ECDF1-disk to
>>>>>>>>>>>>
>>>>>>>>>>> UKI-LT2-IC-HEP-disk (at least all the bits that are contained
>>>>>>>>>
>>>>>>>> within
>>>>>>
>>>>>>> the dirac catalogue):
>>>>>>>>>
>>>>>>>>>> Number of files copied: 1942
>>>>>>>>>>>>
>>>>>>>>>>> Data copied: 7.041 GB
>>>>>>>>> So they should stay accessible no matter what happens to the ECDF
>>>>>>>>> SEs.
>>>>>>>>>
>>>>>>>>>> We just solved the problem of how to remove a VO from the
>>>>>>>>>>>>
>>>>>>>>>>> SE in the
>>>>>>
>>>>>>> storage meeting. It is actually very simple.
>>>>>>>> It doesn't need to have the VOMS information or group information be
>>>>>>>> removed. All that is needed is to change/remove the name of the home
>>>>>>>> directory.
>>>>>>>> So in our case on gridpp09.ecdf.ed.ac.uk, the LSST home directory
>>>>>>>>
>>>>>>> seen > > by
>>>>>>
>>>>>>> DPM was /dpm/ecdf.ed.ac.uk/home/lsst
>>>>>>>> By changing this to /dpm/ecdf.ed.ac.uk/home/test, LSST is no longer
>>>>>>>> reported as supported VO (but we suddenly support the VO TEST...).
>>>>>>>>
>>>>>>>>> Cheers,
>>>>>>>>>>>>
>>>>>>>>>>> Marcus
>>>>>>>> -- > > > > > > The University of Edinburgh is a charitable body,
>>>>>>>>
>>>>>>> registered in
>>>>>>
>>>>>>> Scotland, with registration number SC005336.
>>>>>>>>
>>>>>>>>>
>>>>>>>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>> The University of Edinburgh is a charitable body, registered in
>>>> Scotland, with registration number SC005336.
>>>>
>>>>
>>>
>>>
>>>
>>>
>> --
>> The University of Edinburgh is a charitable body, registered in
>> Scotland, with registration number SC005336.
>>
>
>
>
>
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
|