Hi,
I suscribe David point of view although this means to add a partition for
each experiment, since the lcg tools publish a df output which means that
quotas on the same partition are ignored. And partitioning a disk is not
as flexible as having quotas. At the end I think that this is a problem
with the software that needs to be improved. It doesn't look that
difficult to replace df with quota or to add quota as alternative.
I haven't found any bug opened in savannah about this....
cheers
alessandra
On Tue, 18 Jan 2005, Dr D J Colling wrote:
> Hi Jeremy,
>
> This is clearly a reasonable request provided that the default
> configuration is made to be one where the data from one experiment cannot
> completely stop anything else from happening at the site.
>
> All the best,
> david
>
> On Tue, 18 Jan 2005, Coles, J (Jeremy) wrote:
>
>> Dear All
>>
>> As many of you will already be aware, there have been issues recently with
>> SEs filling up or sites wishing to purge data for other reasons. This matter
>> was discussed at the weekly Grid Operations meeting yesterday and as a
>> result the following request has been made:
>>
>> 1) Please can all ROCs and Tier 1 centres make sure that their sites do
>> *NOT* delete files from SEs, unless specifically requested to do so by a VO
>> manager.
>>
>> Please could you let me know as soon as possible any issues/concerns that
>> you have as a result of this request.
>>
>> A planned TB-SUPPORT meeting for 17th December did not go ahead so I
>> apologise that there has been a gap until now. The next teleconferenece will
>> be Tuesday 25th January at 11am - I hope most of you can take part then. One
>> major topic for the agenda will be migration to SL3. I will send out a full
>> agenda on Friday.
>>
>> Kind regards,
>> Jeremy
>>
>
--
********************************************
* Dr Alessandra Forti *
* Technical Coordinator - NorthGrid Tier2 *
* http://www.hep.man.ac.uk/u/aforti *
********************************************
|