Hi Tom, Catalin,
At a minimum, the documentation should consider the position of a novice
sysadmin who has taken over a site after the incumbent goes to (say)
Patagonia to work as a Gaucho on a cattle ranch.
So we should tell the new fellow what's got to be done, why it has to be
done and how to do it, esp. for something that is a new requirement.
And I might be able to understand it myself if it's that straightforward.
Cheers,
Steve
On 01/26/2015 02:40 PM, Tom Whyntie wrote:
> Hello,
>
> Current CVMFS pages I can find on the wiki:
>
> * https://www.gridpp.ac.uk/wiki/RAL_Tier1_CVMFS - general - does this
> need updating?
>
> * https://www.gridpp.ac.uk/wiki/UK_CVMFS_Deployment - looks pretty
> comprehensive, worth a look
>
> * https://www.gridpp.ac.uk/wiki/CVMFS_Use_Cases_for_GridPP - evolving,
> everyone please feel free to contribute
>
> * https://www.gridpp.ac.uk/wiki/A_quick_guide_to_CVMFS - I think I
> need to update this with the gsi instructions and test commands
> (currently has the web interface instructions which I think is being
> deprecated).
>
> Thanks again everyone, Tom
>
>
>
> On 26 January 2015 at 14:31, Catalin Condurache
> <[log in to unmask]> wrote:
>> Hi Alessandra,
>>
>> The repositories are available for use, both Daniela and Robert were able
>> to upload some data within the appropriate repositories and make it
>> available at WN level.
>> People with appropriate credentials (i.e. role=lcgadmin for each VO) could
>> upload data and as long as the batch farms have correct configuration for
>> the 'gridpp.ac.uk' CVMFS domain (and I think all UK sites have it), the
>> repositories could be mounted.
>>
>> I'll have a look at the existing documentation in the gridpp.ac.uk wiki
>> pages, maybe it needs some updates.
>>
>> Catalin
>>
>>
>> On 26/01/2015 13:01, "Alessandra Forti" <[log in to unmask]> wrote:
>>
>>> Hi Catalin,
>>>
>>> is there any news on this can we mount the repositories for testing now
>>> or do you need to do something else?
>>>
>>> cheers
>>> alessandra
>>>
>>> On 21/01/2015 12:26, Catalin Condurache wrote:
>>>> Hi Robert,
>>>>
>>>> It doesn't exactly work as expected as you uploaded a '-r--------' file
>>>> which cannot be published by CVMFS.
>>>>
>>>> Maybe I should remind everyone, that CVMFS space is a 'public' space
>>>> i.e. readable by _everyone_. Furthermore the repository maintainers have
>>>> to make sure that uploaded files are at least rw-r--r-- and
>>>> (sub)directories are rwxr-xr-x
>>>>
>>>> Cheers,
>>>> Catalin
>>>>
>>>>> -----Original Message-----
>>>>> From: Testbed Support for GridPP member institutes [mailto:TB-
>>>>> [log in to unmask]] On Behalf Of Robert Frank
>>>>> Sent: 21 January 2015 11:26
>>>>> To: [log in to unmask]
>>>>> Subject: Re: CVMFS for regional testbed VOs
>>>>>
>>>>> Hi Catalin,
>>>>>
>>>>> it works for northgrid, thanks.
>>>>>
>>>>> Cheers,
>>>>> Robert
>>>>>
>>>>> On 20/01/15 14:12, Catalin Condurache wrote:
>>>>>> Hi,
>>>>>>
>>>>>> The regional UK VO repositories have been setup as described in the
>>>>> previous email.
>>>>>> For uploading stuff and subsequently maintaining the repositories:
>>>>>>
>>>>>> - the VOMS Role=lcgadmin has been enabled for all these VOs
>>>>>> for GSI
>>>>> access on cvmfs-upload01.gridpp.rl.ac.uk server
>>>>>> - using the above credentials, gsiscp and gsissh connections
>>>>>> will land
>>>>> authorised people in /home/londongridsgm (or /home/scotgridsgm or
>>>>> /home/southgridsgm or /home/northgridsgm) directory; for some
>>>>> technicalities the CVMFS area is /home/londongridsgm/cvmfs_repo (or
>>>>> appropriate) where they should be able to upload (using gsiscp) files /
>>>>> tarballs (that are part of the directory) and then unpack them there.
>>>>>> - when using gsissh / gsiscp, one needs to use port 1975
>>>>>> ('gsissh -p
>>>>> 1975 cvmfs-upload01.gridpp.rl.ac.uk')
>>>>>>
>>>>>> Please try to add some content in the CVMFS areas using this method
>>>>>> and
>>>>> let me know if successful or not.
>>>>>> We¹ll then continue and finalise the last bits of configuration to
>>>>>> make the
>>>>> repositories accessible at WN level.
>>>>>> Regards,
>>>>>> Catalin
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> From: Testbed Support for GridPP member institutes
>>>>>> [mailto:[log in to unmask]] On Behalf Of Alessandra Forti
>>>>>> Sent: 15 January 2015 15:07
>>>>>> To: [log in to unmask]
>>>>>> Subject: Re: CVMFS for regional testbed VOs
>>>>>>
>>>>>> The default role for software maintainance is usually lcgadmin if you
>>>>>> can
>>>>> enable it.
>>>>>> cheers
>>>>>> alessandra
>>>>>>
>>>>>> On 15/01/2015 14:27, Catalin Condurache wrote:
>>>>>> Hi,
>>>>>>
>>>>>> As it was eventually agreed the following CVMFS repositories will be
>>>>>> created
>>>>>>
>>>>>> /cvmfs/londongrid.gridpp.ac.uk
>>>>>> /cvmfs/scotgrid.gridpp.ac.uk
>>>>>> /cvmfs/southgrid.gridpp.ac.uk
>>>>>> /cvmfs/northgrid.gridpp.rl.ac.uk
>>>>>>
>>>>>> About people maintaining the repositories, there is an appetite to
>>>>>> enable
>>>>> VOMS roles for authorization to maintain a repository. I haven¹t tried
>>>>> extensively before (but I have a recipe and have done few tests), but
>>>>> I¹d give
>>>>> it a go. Question is: which is the role for each VO to be enabled for
>>>>> this
>>>>> operation. Probably there is an identical role, but let me know for
>>>>> each VO
>>>>> please.
>>>>>> Regards,
>>>>>> Catalin
>>>>>>
>>>>>>
>>>>>> From: Testbed Support for GridPP member institutes
>>>>>> [mailto:[log in to unmask]] On Behalf Of Peter Gronbech
>>>>>> Sent: 15 January 2015 14:20
>>>>>> To: [log in to unmask]<mailto:TB-
>>>>> [log in to unmask]>
>>>>>> Subject: Re: CVMFS for regional testbed VOs
>>>>>>
>>>>>> Hi Catalin,
>>>>>> We are happy to follow the naming convention you suggest ie
>>>>>> /cvmfs/southgrid.gridpp.ac.uk
>>>>>>
>>>>>> Thanks Pete
>>>>>> --
>>>>>> ----------------------------------------------------------------------
>>>>>> Peter Gronbech GridPP Project Manager Tel No. : 01865 273389
>>>>>>
>>>>>> Department of Particle Physics,
>>>>>> University of Oxford,
>>>>>> Keble Road, Oxford OX1 3RH, UK E-mail :
>>>>>> [log in to unmask]<mailto:[log in to unmask]>
>>>>>> ----------------------------------------------------------------------
>>>>>>
>>>>>> From: Testbed Support for GridPP member institutes
>>>>>> [mailto:[log in to unmask]] On Behalf Of Catalin Condurache
>>>>>> Sent: 13 January 2015 15:24
>>>>>> To: [log in to unmask]<mailto:TB-
>>>>> [log in to unmask]>
>>>>>> Subject: Re: CVMFS for regional testbed VOs
>>>>>>
>>>>>> Hi Daniela,
>>>>>>
>>>>>> Indeed I incline to go for Œgridpp.ac.uk¹ domain which is already
>>>>>> configured
>>>>> within UK and requires the least or none effort.
>>>>>> As for the repo name, would /cvmfs/londongrid.gridpp.ac.uk (or
>>>>> /cvmfs/northgrid.gridpp.ac.uk etc) be satisfactory (my first choice)?
>>>>>> Other options are /cvmfs/vo.scotgrid.gridpp.ac.uk or
>>>>> /cvmfs/london.gridpp.ac.uk, but anyway I¹d like to have consistency
>>>>> among
>>>>> all these regional VOs repos, so need to decide once for all.
>>>>>> Regards,
>>>>>> Catalin
>>>>>>
>>>>>>
>>>>>> From: Testbed Support for GridPP member institutes
>>>>>> [mailto:[log in to unmask]] On Behalf Of Daniela Bauer
>>>>>> Sent: 13 January 2015 14:25
>>>>>> To: [log in to unmask]<mailto:TB-
>>>>> [log in to unmask]>
>>>>>> Subject: Re: CVMFS for regional testbed VOs
>>>>>>
>>>>>> Hi Catalin,
>>>>>> I prefer the gridpp.ac.uk<http://gridpp.ac.uk> simply because these
>>>>>> are
>>>>> truly regional VOs (there won't be any e.g. northgrid in London), but
>>>>> given
>>>>> the small user base and scope of these VOs we should just go ahead with
>>>>> whatever is easiest to setup.
>>>>>> Cheers,
>>>>>> Daniela
>>>>>>
>>>>>> On 13 January 2015 at 10:05, Catalin Condurache
>>>>> <[log in to unmask]<mailto:[log in to unmask]>>
>>>>> wrote:
>>>>>> Hi,
>>>>>>
>>>>>> Following a short email exchange with Alessandra re: CVMFS space for
>>>>> vo.northgrid.ac.uk<http://vo.northgrid.ac.uk>, it emerged the idea to
>>>>> create
>>>>> and maintain a (small) CVMFS infrastructure that would serve mainly
>>>>> the UK
>>>>> regional VOs (vo.southgrid.ac.uk<http://vo.southgrid.ac.uk>
>>>>> vo.northgrid.ac.uk<http://vo.northgrid.ac.uk>
>>>>> vo.scotgrid.ac.uk<http://vo.scotgrid.ac.uk>
>>>>> vo.londongrid.ac.uk<http://vo.londongrid.ac.uk>) separated from the
>>>>> 'egi.eu<http://egi.eu>' CVMFS domain configuration.
>>>>>> One option would be to keep them under the already existing (and
>>>>> configured everywhere in the UK, so no extra work needed!)
>>>>> 'gridpp.ac.uk<http://gridpp.ac.uk>' CVMFS domain (just to decide on the
>>>>> repo names i.e.
>>>>> /cvmfs/northgrid.gridpp.ac.uk<http://northgrid.gridpp.ac.uk> or
>>>>> /cvmfs/north.gridpp.ac.uk<http://north.gridpp.ac.uk> or
>>>>> /cvmfs/vo.northgrid.gridpp.ac.uk<http://vo.northgrid.gridpp.ac.uk> or
>>>>> /cvmfs/vo.north.gridpp.ac.uk<http://vo.north.gridpp.ac.uk>).
>>>>>> But also a new CVMFS domain 'ac.uk<http://ac.uk>' could be defined
>>>>>> (but which would need some extra work for implementation and
>>>>>> configuration at participating sites) with more suggestive repo names
>>>>>> i.e. /cvmfs/vo.northgrid.ac.uk<http://vo.northgrid.ac.uk> or
>>>>>> /cvmfs/northgrid.ac.uk<http://northgrid.ac.uk>
>>>>>>
>>>>>> In any case RAL could act as Stratum-0 and Stratum-1 for these
>>>>>> repositories,
>>>>> but probably we'd need another UK site to volunteer to host a
>>>>> Stratum-1 for
>>>>> failover purposes.
>>>>>> As there are already requests to setup repos for londongrid and
>>>>>> northgrid
>>>>> VOs already, this needs to be discussed and decided asap, via email or
>>>>> maybe
>>>>> today's Ops meeting?
>>>>>> Let me know of your opinions!
>>>>>>
>>>>>> Regards,
>>>>>> Catalin
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Sent from the pit of despair
>>>>>>
>>>>>> -----------------------------------------------------------
>>>>>> [log in to unmask]<mailto:[log in to unmask]>
>>>>>> HEP Group/Physics Dep
>>>>>> Imperial College
>>>>>> London, SW7 2BW
>>>>>> Tel: +44-(0)20-75947810
>>>>>>
>>>>> http://www.hep.ph.ic.ac.uk/~dbauer/<http://www.hep.ph.ic.ac.uk/%7Edba
>>>>> u
>>>>>> er/>
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>>
>>>>>> Respect is a rational process.
>>>>>>
>>> --
>>> Respect is a rational process.
|