Minutes of the ITC-Research Computing Standing Committee

Meeting on November 25, 2002

 

Members:  Alice, Bill, Brian, Dawn, David, Ed, Hamp, Kathy, Jim, Mark S., Martha, Michael, Robin, Sue Ellen, Steve, Terry, Tim S., Tom S.

 

Attending: Alice, David, Dawn, Ed, Hamp, Kathy, Jim, Mark S., Robin, Steve, Tim T.

 

Chair:  Tim T.

Recorder:  Alice

 

I.  Discussion Topics:

 

1.  Final check of final quote from Aspen on new cluster:  Tim has confirmed with Aspen that the floppy drives are built into the chassis; however it’s possible that the nodes do not have CD-ROM drives (although the head does) – this could cause a problem with the way we usually do the initial configuration of each node.  After checking the Aspen Systems website for more information (inconclusively), Tim will just check with Alan during his conference call this morning. (Note:  Tim, Dawn, and Hamp are going to talk with Alan this morning about “support issues” such as our method of arranging for replacement parts, getting correct replacement parts, longer-term availability of parts, etc.)  The quote for the AMD-related parts is fine as-is.  Aspen is willing to give us a credit for the cluster management software that we are not using on the current cluster.

 

2.  Replacement for the Unix Lab “status” web page:  Steve has been working on a replacement for the current Unix Lab status page – see http://www.itc.virginia.edu/unix/gmond -- the new page uses Ganglia and has an option to view the status of the new cluster.  In reviewing the new page, there was a suggestion that perhaps we should change the name “Aspen Linux Cluster” to something that is not specifically tied to the vendor.

 

3.  Timeline to move AIX SMP box to PBSPro:  it was concluded that the safest way to go is to use a separate AIX box (Athena) for a head node; Hamp thinks these changes could be done between the Fall and Spring semesters (i.e. sometime between 12/16/02 and 1/15/03).

 

4.  SEAS J40/J50 Migration:  Tim and Hamp have reviewed Mitch’s draft letter which will go out soon.  It will recommend that users move off these systems by June 2003 (although the systems will remain up, but out of maintenance, after that date).  From the process accounting logs, Hamp can tell that there are fewer active users than actual account holders on these systems.

 

 

Next regular meeting:  Monday, December 16, 2002 at 10:30 AM in 117 Astronomy