From: Albert Lazzarini Date: Tue Feb 26, 2002 12:22:25 US/Pacific To: lsccompcomm@ligo.caltech.edu Cc: Albert Lazzarini Subject: Minutes of LSC COMPUTING COMMITTEE meeting of 2002.02.12 Minutes recorded by A. Wiseman, edited by A. Lazzarini ( where noted by AL: ) 1. Near term allocation of LDAS resources. It has been proposed that the immediate responsibility of resource allocation should be placed with the upper limit chairs, who know what search algorithms are being implemented within their groups. One action that should come of the LSC E7 review meeting on 20 February is some insight in relative computational needs among the groups. AGW: I have no notes on addressing this point. Comment added by AGW: A poll of the UL Chairs has been conducted. They beleive they have sufficient resources to do the work. I have my doubts. AL: The issue of how the searches scale is under study and until we know what this scaling is, we have uncertainties in requirements. Early scaling studies pointed to architectural improvements needed in the construction of some codes; we do not know where the "hardware brick wall" is yet. 2. Teragrid has requested input from LIGO regarding how it (the collaboration) might want to use the teragrid facility at CACR. The request requires a response in the form of a proposal, because resource use is awarded on a peer-review basis that is performed by an advisory committee to Teragrid. I sent around a notice of this from Roy Williams (it is in the mail archive: see [Lsccompcomm] Fwd: Williams Archive: Tue Jan 29, 2002 05:05:34: Subject: Invitation to use the Teragrid ). We need to indicate to Teragrid our interest in use of this facility before the end of February. Albert posed the quesiton: Should we partcipate? TN: Answered Why Not? LSF: Reason not to: we are tasked saturated. PB: suggested that the commitment might not be so great. We can take a wrapped stand-alone LAL code as an executable. LSF: suggested that the notion of simply "wrapping LAL code" should be revisited in more detail. AGW concurred that some additional specification might be necessary. Kent mentioned some efforts to port the stand alone wrapper to the grid. Bottom Line: General agreement: we should be willing to participate, but should agree to a large commitment of resources now. Albert said he would convey this to Roy Williams. 3. Planning for the FNAL meeting. Not addressed ? AL: Reserved for next week's discussion. 4. Design of new MOU attachments for Tier II centers. 4a. One additional agenda item: I would like a brief status from Patrick, Sam, Tom on their thinking about what the new MOU attachment for institutions with significant hardware resources should contain. A number of random thought were expressed, but I am not sure any clear course of action emerged. Patrick: We have distributed, heterogenous facilities running LDAS; therefore we need a proceedure for agreesing on upgrades. TN: The ordinal numbers are one - two - many. TN: We are all one laboratory. AGW: We are all one collaboration. Need a "plan" from which an MOU should emerge. The plan should address the following (much like iVDGL): -- Hardware (Procurement and allocation) -- Operating systems (version control) -- Infrastructure -- Applications -- Data distribution rules Need to decide how decissions are to be made. Lots of rambling talk. ------------------------------------- LIGO Laboratory at Caltech Data and Computing M/C 18-34 Room 609 Millikan Library 1200 E. California Blvd. Pasadena, CA 91125 Office: +1.626.395.8444 Facsimile: +1.626.304.9834 -------------------------------------