John,

I've marked the nodes offline so no new jobs can run once the current ones
have completed. When the running jobs do have completed I will put the
reservations in. It's difficult to predict how long individual jobs take to
complete because very often users request more wall time than what they will
eventually need.

Jens

-----Original Message-----
From: John Karro [mailto:[log in to unmask]]
Sent: Monday, July 21, 2014 12:09 AM
To: Mueller, Jens Dr.
Cc: Research Computing Support
Subject: Re: Large memory nodes

Jens,

    Yes, we would like to reserve two large memory nodes.  I'm not sure how
kong we need them -- can you give them to us for four days?

-- John Karro
Sent from my iPad, hence subjected to iPad auto-"correct" induced errors.
Which I just refuse to fight with.

> On Jul 18, 2014, at 7:03 PM, "Mueller, Jens Dr." <[log in to unmask]>
> wrote:
>
> Hi John,
>
> Yes, he should have access. The cluster is busy right now but I can
> put in a reservation if he needs the nodes for a extended period of
> time. Jobs are running there right now so we can't kick users off
> immediately. The resource request needs to have:  nodes=1:ppn=8:m128
>
> On a separate note: Robin is leaving Miami and Brad Koby is taking over.
> Jens
>
>> On Fri, Jul 18, 2014 at 5:18 PM, Karro, John <[log in to unmask]> wrote:
>> My graduate student, liux17, needs to start running tests on the
>> large memory nodes on redhawk.  Does he have access, or does he need
>> to be given permission?
>>
>> Also: can you please remind us of the settings needed to direct a job
>> to them?
>>
>>    John
>> ---------------------------------------------------------------------
>> -------------------------
>> Dr. John Karro, Associate Professor
>> Department of Computer Science and Software Engineering
>> Affiliate: Department of Microbiology, Department of Statistics
>> Office: Benton 205D, Miami University, Oxford, Ohio
>> ---------------------------------------------------------------------
>> -------------------------