Here is information about the setup at Argonne. A few requests:
1) Could Argonne account holders check if they can log in to these
machines? I can log in fine, but want to confirm if we have off-site access.
2) Please send me a list of the software you need for the parts of the demo
you are in charge of; both things that could be shared (such as libc
version) and things that are fusiondemo-specific (such as IDL, MDSplus,
etc) If you are not sure send me mail also (Qian, we'll just have to ask
them to install whatever version works for you eventually). Note that this
information is for the desksides only, the clusters are coming later.
3) Let me know if you have any questions/suggestions/requests concerning
this setup.
>From: "William E. Allcock" <allcock@mcs.anl.gov>
>To: <dsl@mcs.anl.gov>, "Todd Tannenbaum" <tannenba@cs.wisc.edu>,
> "Ann Chervenak" <annc@ISI.EDU>, "Carl Kesselman" <carl@isi.edu>,
> "Ewa Deelman" <deelman@ISI.EDU>, "June-Sup Lee" <june@isi.edu>,
> "Karl Czajkowski" <karlcz@isi.edu>, "Laura Pearlman" <laura@isi.edu>,
> "Mei-Hui Su" <mei@isi.edu>, "Steven Fitzgerald" <steve@ecs.csun.edu>
>Cc: "Miron Livny" <livny@cs.wisc.edu>
>Subject: SC infrastructure and Demo Needs
>Date: Wed, 17 Oct 2001 14:44:18 -0500
>X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0)
>Importance: Normal
>Sender: owner-dsl@mcs.anl.gov
>
>The booth infrastructure is operational (details below). At this point,
>everyone needs to begin to migrate their demos over to this infrastructure.
>We now need to figure out what should be installed globally in /soft, and
>what should be in your home directories. So, with that in mind, I need
>everyone who is responsible for a demo to send me a list of EXACTLY
>(including version numbers) what software you need, and I need this ASAP
>(preferably NOW). This is so that if we see multiple people needing the
>same software we can designate one person to install it in /soft and then
>point other people at it. Some of the important issues:
>
>- What version of Globus are you using and which parts
> - If we end up having to run two multiple versions of globus, some of the
> services will have to be on non-standard ports.
>- If you have any dependencies on specific glibc versions, etc
>- What versions of Perl, Python, Java, ad nauseum do you need.
>
>The general idea is that you will be responsible for installing the vast
>majority of your software. If we see that multiple demos will be using the
>same software (like Globus), we will either designate someone to install it
>in /soft or Charles will do it, particularly if it requires root access.
>
>
>
>Following is the status of the SC infrastructure:
>- The MCS home server (sc2001server.mcs.anl.gov) is in place
>- Networking is in place (currently 100 Mbs, will be GigE shortly)
>- hostnames are as they will be at SC (you are going to just love the names
>they
> picked)
>- you can log in with your MCS account names
>- Five of our six desksides are up
>- The GeForce3 cards are not installed yet, but will be by SC (for you
>Graphics
> folks), we could not get Gladiac 920s so we are using Visiontek GeForce3.
>- the clusters are not quite there yet. Hopefully, the clusters will be up
> early next week
>- No software, other than OS (RedHat 7.1), has been installed yet
>- We will take care of getting host certs installed.
>
>Directory Structure:
>- There will be home for each account served of the MCS (booth wide) server.
> - SW specific to your demo should go here
> - Data sets specific to your demo should go here.
>- There will be a shared /soft served off one of our desksides (globus wide)
> - Shared software should be here (Globus, iperf, perhaps Java if a number
>of
> people can use the same version)
>- There is a minimum of 36GB of local disk on each machine
> - It is recommended that this be used as dynamic scratch, so that demos
> can be run from any machine.
> - However, if you need data locally for performance reasons, contact
> Charles or myself and we can discuss this on a case by case basis.
>
>For now, here is how we get access to the machines:
>a031.r352.showfloor.sc2001.org
>a033.r352.showfloor.sc2001.org
>a034.r352.showfloor.sc2001.org
>a035.r352.showfloor.sc2001.org
>a036.r352.showfloor.sc2001.org
>
>Here's how you get to them:
>First, ssh to sc2001server.mcs.anl.gov. From there, you can drop the
>".r352.showfloor.sc2001.org" part of the name, and just ssh to a031,
>a033, etc. For now, these machines are only accessible through
>sc2001server. They will always be accessible this way, even if DNS on the
>show floor goes down. However, once DNS is up you will be able to log into
>the machines directly.
>
>--------------------------------------------
>William (Bill) E. Allcock
>Argonne National Laboratory
>Bldg. 221, Office C-115A
>9700 S. Cass Ave
>Argonne, IL 60439-4844
>E-Mail: allcock@mcs.anl.gov
>Office: 630-252-7573
>Mobile: 630-247-1647
__________________________
Dr. Kate Keahey
Math & Computer Science Div.
Argonne National Laboratory
Argonne, IL 60439, USA
(630) 252-1673
===============================================================================
This message was sent to the SciDAC National Fusion Collaboratory (NFC)
workers list nfc-sc01. Visit the Collaboratory at
<http://www.fusiongrid.org/>.
To unsubscribe from this list, please send a message to
majordomo@fusion.gat.com with the following text in the *body* of the
message: unsubscribe nfc-sc01
David P. Schissel: <schissel@fusion.gat.com> <http://fusion.gat.com/~schissel/>
This archive was generated by hypermail 2.1.1 : Fri Feb 08 2002 - 13:34:29 PST