CMS are modifying phedex to use FTS as well.
However, it must be said that FTS is not quite as transparent as we'd
like yet. The concurrent tests show real problems in recovering from
transient failures. This should improve in later versions of FTS.
Of course, you can also have local access to files through rfio and
dcap, which will stall if the SE is rebooted.
On the whole though, it would be far more disruptive to take the
whole site down just for a 5 min reboot. With current job success
rates on the grid, a rare transient job failure from an SE reboot is
not terribly significant.
Cheers
Graeme
On 23 Mar 2006, at 12:33, Greig A Cowan wrote:
>> Are the sort of transfers done by VO production or analysis jobs
>> using FTS
>> though?
>
> atlas, alice and lhcb are all interfacing their data management
> software
> with FTS. cms have a system called Phedex which provides reliable file
> transfer. Tim Barrass at Bristol can tell you more about it if you're
> interested.
>
> Cheers,
> Greig
>
>
>>
>> Cheers,
>> Simon
>>
>>
>> On Thu, 23 Mar 2006, Greig A Cowan wrote:
>>
>>> Hi Simon,
>>>
>>> Since FTS provides reliable file transfer, you should be able to
>>> reboot
>>> your SE anytime, even during a transfer. Since SRM_PUTDONE will
>>> not be
>>> returned, the transfer will not be marked as successful and FTS will
>>> try again. It will try this 3 times.
>>>
>>> Cheers,
>>> Greig
>>>
>>>
>>>
>>> On Thu, 23 Mar 2006, Simon George wrote:
>>>
>>>> Hi,
>>>>
>>>> I've seen procedures discussed for draining queues so that
>>>> worker nodes
>>>> can be rebooted, but what is the procedure to reboot one of the
>>>> storage
>>>> nodes (in my case a DPM pool node)?
>>>>
>>>> In theory access could come from anywhere not just the local
>>>> jobs, so
>>>> draining queues seems to be irrelevant. I don't know how to tell if
>>>> someone is accessing the SE so I can pick a time when is is not
>>>> being read
>>>> from or written to, if such a time exists. The reboot should
>>>> only take ~5
>>>> mins. Any suggestions?
>>>>
>>>> Cheers,
>>>> Simon
>>>>
>>>
>>> --
>>> ====================================================================
>>> ====
>>> Dr Greig A Cowan http://www.ph.ed.ac.uk/
>>> ~gcowan1
>>> School of Physics, University of Edinburgh, James Clerk Maxwell
>>> Building
>>>
>>> TIER-2 STORAGE SUPPORT PAGES: http://wiki.gridpp.ac.uk/wiki/
>>> Grid_Storage
>>> ====================================================================
>>> ====
>>>
>>
>
> --
> ======================================================================
> ==
> Dr Greig A Cowan http://www.ph.ed.ac.uk/
> ~gcowan1
> School of Physics, University of Edinburgh, James Clerk Maxwell
> Building
>
> TIER-2 STORAGE SUPPORT PAGES: http://wiki.gridpp.ac.uk/wiki/
> Grid_Storage
> ======================================================================
> ==
--
Dr Graeme Stewart - http://wiki.gridpp.ac.uk/wiki/User:Graeme_stewart
GridPP DM Wiki - http://wiki.gridpp.ac.uk/wiki/Data_Management
|