On 21 December 2010 11:32, John Gordon <[log in to unmask]> wrote:
> I agree it looks the way forward but still a risk until there is a production quality service at CERN.
Indeed; the current position of the storage group on CVMFS is
precisely that: "It looks brilliant, but we don't advise switching
until it's officially open."
> I think RAL are running a mirror that can be manuually swung in over Xmas if required but be aware that if anything happens the experiments will expect an immediate fallback. T1s with callout can probably do this but I doubt if you want to sign up to it.
>
> It is still not obvious who will support all the sites trying to change. I guess we have enough expertise in the UK now but the developers don't have the effort to provide deployment and operations support.
>
I think Chris has provided some perspective on how he did stuff at
QMUL, so we do have some T-2 expertise here. I think we'll see how it
goes.
Sam
> John
>
> ________________________________________
> From: Testbed Support for GridPP member institutes [[log in to unmask]] on behalf of Sam Skipsey [[log in to unmask]]
> Sent: 21 December 2010 11:15
> To: [log in to unmask]
> Subject: Re: Deployment team & sites meeting
>
> On 21 December 2010 11:13, John Gordon <[log in to unmask]> wrote:
>> Has widespread deployment of CVMFS started?
>
> No officially, no.
>
>> I thought this was still at the stage of T1 testing. Aren't you jumping the gun before there is a production service at the CERN end?
>>
>
> I think that a lot of sites feel that CVMFS for software areas is so
> much better than the current unscalable approach that they're happy to
> move even when it's in testing. QMUL hasn't had any problems that I am
> aware of, and they've been using it for a while.
>
> Sam
>
>> John
>>
>> ________________________________________
>> From: Testbed Support for GridPP member institutes [[log in to unmask]] on behalf of Matt Doidge [[log in to unmask]]
>> Sent: 21 December 2010 10:58
>> To: [log in to unmask]
>> Subject: Re: Deployment team & sites meeting
>>
>> Hello,
>> Forgot my laptop today, so unless I find another way to join I too will
>> have to send my apologies.
>>
>> Quick Lancaster update in case I don't make it:
>> -We're in the middle of installing glite-APEL, the VM's instantiated,
>> certificate installed. Just need to dump the old database and run yaim.
>> -CREAM CE passing SAM tests, will be doing proper work for atlas once we
>> get the software installed. We will move to CVMFS in the New Year.
>> -We have someone on call throughout the Christmas period so any problems
>> can be dealt with.
>>
>> Have a good Christmas and a problem-free New Year everyone!
>>
>> Matt
>>
>> Sam Skipsey wrote:
>>> Also, for Wahid, I should give apologies. I doubt he's going to be
>>> around for the meeting (he gave prior warning to the Storage Group
>>> that he'd not make our meeting on Wednesday).
>>>
>>> Sam
>>>
>>> On 20 December 2010 21:19, J Coles <[log in to unmask]> wrote:
>>>> Dear All
>>>>
>>>> There will be a meeting Tuesday 21st in EVO at 11am. The agenda is here: http://indico.cern.ch/conferenceDisplay.py?confId=113880. Pete has kindly agreed to chair as I will not be able to join.
>>>>
>>>> For minutes the order is: Mingchao -> Alessandra (away) -> Derek -> Jeremy (away) -> Kashif -> Sam -> Wahid -> Stuart.
>>>>
>>>> Kind regards,
>>>> Jeremy
>>>>
|