Minutes for the ITC-Research Computing Standing Committee Meeting

October 31, 2005 at 10:30AM

 (in ITC-2015 Ivy Road, Room 102)

 

Members: Alice, Bill, Dawn, Ed, Hamp, Kathy G., Katherine H., Jim, Mark S., Martha, Nancy K, Robin, Sue Ellen, Steve, Terry, Tim S., Joe S., Tom S.

 

Attending:  Alice, Bill, Ed, Katherine H., Jim, Mark S., Martha, Robin, Terry, Tim S., Joe S.

Chair: Tim T. Recorder: Alice Howard

 

I. Corrections to minutes from last meeting on August 29, 2005?  No.

 

Click here to see the agenda from this meeting.

 

II. Ongoing Discussion Topics:

 

1.  Cluster projects:

·    “Condo” Cluster (aka cedar):

o        FIVE researchers participating: Leo Zhigilei (lz2n) – 8 nodes w/ ETF $; Jeff Shabanowitz/Dina Bai (Chemistry)- 8 nodes; Phil Parrish/Rich Gregory – 4 nodes; Matt Neurock – 4 nodes; new researcher, Avik Ghosh (av7rq), EE, 10 nodes.

o        Outstanding issues:

§          Is AVAKI Gateway node on Cedar?  On the test cluster – need to ping Steve re testing on the test cluster.

§         PBS scheduler groups for “condo purchasers” – Katherine getting list of users.

§         PBSPro priority queuing for paying participants and usage reporting.

§         Power replacement – still waiting for the plug converters – will give folks a least a weeks notice before scheduling a shutdown.  Meanwhile all bad PDUs are out and power has been stable the past month.

·    Aspen & Birch cluster:  Upgrade done.  Any issues or concerns?

o        Maintenance will be gone on Birch in December, 2005 – any plans to renew?  Under discussion.

·    All the Linux clusters: At August meeting agreed to create script/alias for most frequently used password commands (e.g., passwd) to invoke message that users need to change password on blue.unix not on cluster front-end.  Please complete by November 11.

o        Tim will email Hamp and Bill about getting this done.

·    Oak Cluster:  (Outstanding from June) Adding the modules software primarily to make it seamless for IMSL and Java users to move from aspen to oak and make it easier for users to access different versions of the Solaris compilers.  Complete by December 16.

o        Done.

·    Teak cluster:  OK.

 

 

2.  /longtmp

·    Script to automate checking usage and reminding users.  (Hamp and TimT to collaborate on this.)  Complete by December 16.

o        No change in status.

 

 

3.  blue.unix upgrade of nodes to new 64-bit nodes and new version of OS. ITC-Transitions CDP is tracking as well.  Timeline?  We need to test software carefully.

·    Rollout timeline?  Logging in to holmes is still slow – not going anywhere the way it is now.

·    Will “gather” command/tools still work?

·    Are there other software packages that were node specific (e.g., Lisrel was) that we need to take into account?

4.  Ganglia web pages:  Need to add cedar.itc.

·    Can we get  teak and oak clusters separated out from the general research computing cluster so we can see these nodes easily?

·    Is it possible to add athena.itc (SMP)? 

·    TimT will ping SteveL.

 

5.  Andew Grimshaw’s request for Linux front-end to campus grid project – Hamp was going to follow-up on this.

·    Wait to hear from Hamp – we probably have told Andrew that we will provide a machine.

 

III. New Business:

6.  Marty Humphrey’s grid computing project  (Katherine)

·    Has own grid project that is globus based – and sort of works.  Seems like it might be useful for users to submit jobs via the web – also help with their file transfers and management. 

·    Katherine would like to see some enhancements to the web service – and we should be working with him on NetBadge and the uPortal framework.

 

7.  Phil Parrish leading UVa’s participation in responding to NSF’s Request for Proposals for next generation/iteration of Supercomputing Centers.  UVa participating via ORNL and its consortium schools and likely one or two others outside consortium.

·    Does UVa want to get involved?  If so, how much? Infrastructure?  Right now Mitch R. is “holding the ball” and Oak Ridge is pulling this together.

 

8.  Next “Condo” Cluster – what possible specs to advertise:  dual core Opteron chips?  2GB memory.  Price range in $2,000 to $2,500/node?  4 node minimum order.

·    Don’t necessarily assume dual core.  We should decide based on price and performance.  Our aim is to purchase a system around the same size as cedar.

 

 

Next Scheduled Meeting of this group: Monday, December 19, 10:30 AM in ITC-2015 Ivy Road, Room #102.

 

Next Scheduled Meeting of ITC ResComp Management team: Monday, November 28, 10:30 AM in ITC-2015 Ivy Road, Room #220.  (Note:  This meeting was postponed and rolled into December 16th meeting.)

 

  

 <--- Go to: ITC Research Computing Standing Committee Home Page