Dear All
Clearly lots to follow up from this particular biomed episode.
Due to the concerns I raised ticket 28530 this afternoon with a "top priority" setting. Flavia Donno (used to be EIS now working on many things!) has given an initial response indicating that the jdl did correctly specify the memory requirements and for the sites in question to check their configuration. Heinz (user in question - see other thread "Heinz' challenge") was copied on the ticket but has yet to respond.
In parallel (given the urgency) I mailed the VO manager (Yannick Legre) and the VO experts (as listed in the VO card - Johan Montagnat and Christophe Blanchet). Johan replied within a few hours:
"I have no idea what the "gnfs-lasieve" program is suppose to do. Why is it
considered "not appropriate for the biomed VO"? Another ticket was sent
yesterday mentionning that Heinz Stockinger was running RSA cracking
code and baned from Oxford site. Is this this "gnfs-lasieve" code?
I have contact Heinz directly today asking him to clarify his activity. If
there is no satisfying answer from him I will ask that he is banned from
the VO."
Yannick also replied shortly afterwards:
"As far as I know Heinz is working for the Swiss Institute of Bio-informatic and I'm not sure he's trying to crack anything...
As Johan, I'm interested to know what it is not appropriate for biomed? (please see http://en.wikipedia.org/wiki/General_number_field_sieve)"
In response I have highlighted the statement in the VO use policy (again linked from the VO ID card) that for biomed "... The goal of the VO is to foster life sciences collaboration taking advantage of the Grid." Clearly if this is RSA cracking it is not appropriate use - though the policy quoted here could do with less ambiguity about what constitutes fostering collaboration! I've not seen any special request that the biomed remit be extended for a special challenge.
Wider than biomed I can see several general points that need to be addressed about types of jobs that VOs may run and how well communication works through GGUS and elsewhere. There are now metrics on ticket response times and this ought to be used to encourage users as well as sites through service agreements/descriptions. I'll follow up on this idea. As for checking that VOs are not misusing their access this is more difficult and will require some thought. [Aside: The gridpp VO (which I manage) for example is getting quite large with quite diverse communities joining it. We have actively encouraged its use to get other/new users on the grid - in fact that is why it was set up. Today's events might be a good prompt to make faster progress with the regional VOs].
Regards,
Jeremy
> -----Original Message-----
> From: Testbed Support for GridPP member institutes [mailto:TB-
> [log in to unmask]] On Behalf Of Graeme Stewart
> Sent: 01 November 2007 00:22
> To: [log in to unmask]
> Subject: Re: biomed jobs
>
> See https://gus.fzk.de/ws/ticket_info.php?ticket=28493, which I sent
> 24 hours ago (and as I am TPM I assigned it to biomed VO immediately).
>
> No response.
>
> g
>
>
> On 1 Nov 2007, at 00:14, Gordon, JC (John) wrote:
>
> > With reference to the other discussions, I don't think we have
> > given this route much time to progress. The ticket was only
> > assigned to biomed at 1637 GMT (1737 in France) and it is not
> > flagged as a VO-specific problem.
> >
> > John
> >
> >> -----Original Message-----
> >> From: Testbed Support for GridPP member institutes
> >> [mailto:[log in to unmask]] On Behalf Of Ma, M (Mingchao)
> >> Sent: 31 October 2007 23:04
> >> To: [log in to unmask]
> >> Subject: Re: biomed jobs
> >>
> >> A ticket has been opened by Jeremy
> >> (https://gus.fzk.de/pages/ticket_details.php?ticket=28530)
> >> and the issue has also been reported to the related VO
> >> manager. We are waiting for their response. A grid-wide
> >> security operation via OSCT will be taken if we do not have
> >> appropriate explanation.
> >>
> >> Mingchao
> >>
> >>
> >>> -----Original Message-----
> >>> From: Testbed Support for GridPP member institutes
> >>> [mailto:[log in to unmask]] On Behalf Of Graeme Stewart
> >>> Sent: 31 October 2007 16:08
> >>> To: [log in to unmask]
> >>> Subject: Re: biomed jobs
> >>>
> >>> Have a look at the outputs I posted as part of my ticket,
> >>>
> >>> https://gus.fzk.de/pages/windownload.php?attid=ATT2653
> >>>
> >>> There are files marked rsa768, so I agree with Kostas. It's
> >> the $50k
> >>> prize. (I can send the siever-worker.pl script if anyone
> >> wants.) If I
> >>> hadn't been convinced I wouldn't have killed the jobs.
> >>>
> >>> If he wins, Glasgow want £2k towards running costs. Other T2s might
> >>> like to estimate their costs.
> >>>
> >>> g
> >>>
> >>> PS. GNFS might have something to do with blood, but you
> >> can't lasieve
> >>> blood vessels (see
> >> http://www.math.ttu.edu/~cmonico/software/ggnfs/).
> >>>
> >>> PPS. If he lost his credential then the it started to be
> >> abused around
> >>> 2007-09-19 by our records.
> >>>
> >>>
> >>> On 31 Oct 2007, at 15:36, Ma, M (Mingchao) wrote:
> >>>
> >>>> Yes, agree. But first, we need to confirm if it is indeed
> >> an abuse.
> >>>>
> >>>> Mingchao
> >>>>
> >>>> ________________________________
> >>>>
> >>>> From: Kelsey, DP (David)
> >>>> Sent: Wed 31/10/2007 15:01
> >>>> To: 'Testbed Support for GridPP member institutes'
> >>>> Cc: Ma, M (Mingchao)
> >>>> Subject: RE: biomed jobs
> >>>>
> >>>>
> >>>>
> >>>> Dear all,
> >>>>
> >>>> The User AUP says...
> >>>>
> >>>> You shall only use the GRID to perform work, or transmit or store
> >>>> data consistent with the stated goals and policies of the VO of
> >>>> which you are a member and in compliance with these
> >>> conditions of use.
> >>>>
> >>>> So... It then depends of course on what the VO AUP says.
> >>>>
> >>>> If work is not consistent with the AUP then it should be
> >>> reported -
> >>>> not sure we have a clear procedure defined for this - I suggest
> >>>> people should report to the UK/I Security Officer (cc'd). This
> >>>> should then be reported to the appropriate VO manager
> >> and/or Grid-
> >>>> wide security operations.
> >>>>
> >>>> And please... Don't assume that the credential under which
> >>> the work
> >>>> is being executed is the responsible party - this credential may
> >>>> have been compromised.
> >>>>
> >>>> Dave
> >>>>
> >>>>
> >>>>
> >>>> ------------------------------------------------
> >>>> Dr David Kelsey
> >>>> Particle Physics Department
> >>>> Rutherford Appleton Laboratory
> >>>> Chilton, DIDCOT, OX11 0QX, UK
> >>>>
> >>>> e-mail: [log in to unmask]
> >>>> Tel: [+44](0)1235 445746 (direct)
> >>>> Fax: [+44](0)1235 446733
> >>>> ------------------------------------------------
> >>>>
> >>>>
> >>>>> -----Original Message-----
> >>>>> From: Testbed Support for GridPP member institutes
> >>>>> [mailto:[log in to unmask]] On Behalf Of Cornwall,
> >>> LA (Linda)
> >>>>> Sent: 31 October 2007 14:32
> >>>>> To: [log in to unmask]
> >>>>> Subject: Re: biomed jobs
> >>>>>
> >>>>>> -----Original Message-----
> >>>>>> From: Testbed Support for GridPP member institutes [mailto:TB-
> >>>>>> [log in to unmask]] On Behalf Of Kostas Georgiou
> >>>>>> Sent: 30 October 2007 17:18
> >>>>>> To: [log in to unmask]
> >>>>>> Subject: Re: biomed jobs
> >>>>>>
> >>>>>> On Tue, Oct 30, 2007 at 05:02:20PM +0000, Kostas Georgiou wrote:
> >>>>>>
> >>>>>>> Hi,
> >>>>>>>
> >>>>>>> Does anyone know what the gnfs-lasieve program that biomed is
> >>>>> running?
> >>>>>>> With a quick look I found
> >>>>>> http://www.math.ttu.edu/~cmonico/software/ggnfs/
> >>>>>>> which suggests that they are running some factorization code
> >>>>> probably
> >>>>>>> trying to break the RSA-768 challenge (it has a $50K reward
> >>>>> afterall).
> >>>>>>
> >>>>>> A closer look shows that indeed the user is trying for the $50K
> >>>>> prize!!!
> >>>>>
> >>>>>
> >>>>> If this is true it must be against policy. I suspect
> >> some sort of
> >>>>> action needs to be taken. Maybe Dave can comment?
> >>>>>
> >>>>> Linda.
> >>>>>
> >>>>>>
> >>>>>> Cheers,
> >>>>>> Kostas
> >>>>>>
> >>>>>> PS> Should we bill the user or the VO for the cpu usage?
> >>>>>
> >>>
> >>> --
> >>> Dr Graeme Stewart -
> >> http://wiki.gridpp.ac.uk/wiki/User:Graeme_stewart
> >>> ScotGrid - http://www.scotgrid.ac.uk/ http://scotgrid.blogspot.com/
> >>>
> >>
>
> --
> Dr Graeme Stewart - http://wiki.gridpp.ac.uk/wiki/User:Graeme_stewart
> ScotGrid - http://www.scotgrid.ac.uk/ http://scotgrid.blogspot.com/
|