Hi Daniela,
You'll be pleased/saddened to know that Atlas also pull stuff like this
reasonably often - usually it's users not sanitising their analysis jobs
before grid submission.
The standing word from atlas uk is if we see user jobs that are using
more cores then they should we're free to kill them. I presume they mean
the jobs themselves, and not the submitters...
Cheers,
Matt
On 20/06/16 14:40, Daniela Bauer wrote:
> I don't know if this (I?) can be helped, but at this stage I'd settle
> for sympathy:
> CMS multicore jobs at Imperial:
> https://ggus.eu/index.php?mode=ticket_info&ticket_id=119617
>
> Is it just us ? Or do other people just not care (e.g. we got pulled
> up for killing single core jobs running in 8 core slots) ? Does Atlas
> pull stunts like this ?
>
> Cheers,
> Daniela
>
> On 20 June 2016 at 14:34, Terry Froy <[log in to unmask]> wrote:
>> On 20/06/2016, 14:27, "Testbed Support for GridPP member institutes on behalf of Matt Doidge" <[log in to unmask] on behalf of [log in to unmask]> wrote:
>>
>>> <snip>
>>> QMUL
>>> https://ggus.eu/?mode=ticket_info&ticket_id=122193 (16/6)
>>> Multiple DNS entries for the QM (dual stacked) perfsonar hosts are
>>> causing intermittent test failures, Duncan points to documentation that
>>> asks that perfsonar hosts only have one hostname. Assigned (16/6)
>>
>> Hi Matt,
>>
>> The underlying issue was fixed last week – there was an additional PTR record in DNS.
>>
>> I am unable to resolve the ticket in GGUS though.
>>
>> Regards,
>> Terry
>> --
>> Terry Froy
>> Cluster Systems Manager, Particle Physics
>> Queen Mary, University of London
>> Tel: +44 (0)207 882 6560
>> E-mail: [log in to unmask]
>>
>
>
>
|