From: lsst-dmlt-bounces@lsstcorp.org on behalf of Mario Juric Sent: Tuesday, December 20, 2016 5:09 PM To: Cc: Victor Krabbendam Subject: [LSST-DMLT] DRAFT: DMLT Meeting Minutes 2016-12-19 [[ Everyone: PLEASE READ & DOUBLE-CHECK/RESPOND WHERE NEEDED ]] Hi everyone, Below is what I captured from the DMLT meeting yesterday. Please amend as necessary and I’ll send out the final version of the minutes. * As of this week, Jacek is on vacation (and SLAC in general is closing for the year after COB tomorrow). Jacek will be returning for a few days in January, but and those will be his last with us before moving to Teradata. We’ll miss him dearly, and hope he has a good vacation and a great start at TD making big data seem small! * After Jacek’s departure, Victor will take a more direct role in leading the final phases of the replan. He’s already joined the TCAM meeting last Friday, and a summary of the conclusions has been circulated. This will only intensify in January and February, until Wil assumes his position. * Our primary replan goal through the end of the year is to assemble an internally self-consistent plan. To get there, all teams have been busy understanding the design, capturing work packages, and (most recently) identifying interfaces and inter-team milestones. We’re in a position take a first stab at assembling that integrated plan. - By Friday: MJ will take the activities/milestones from all teams and attempt assemble them into an integrated plan (in Omniplan, for now). Will begin with Princeton and UW activities, then add others. * Here’s what I captured people as working on this week (please correct anything I got wrong). I also added the “expect to have by Friday” line — can all the TCAMs fill it out, to make sure I have it captured correctly?: Princeton (John): * Making final updates to the work packages based on feedback from Jim * Iterating on milestones with other groups * Begun discussing possible descope options with Jim, Robert, Zeljko * Activities are in OmniPlan, milestones are in the milestone spreadsheet * Will have by Friday: ***** PLEASE FILL OUT ****** Tucson (Frossie): * There are a few blockers to discuss with the Pipelines group, will do that * Writing an architecture document for the QC system * SQuaRE activities are in the Google doc spreadsheet * Will have by Friday: ***** PLEASE FILL OUT ****** UW (Simon): * Plan is in a PM tool (apologies for not catching the name) * Milestones are in the Google Doc spreadsheet * Should have a version ready by the end of the week * Will have by Friday: ***** PLEASE FILL OUT ****** SLAC (Fritz): * Activities in the Google Doc spreadsheet * Milestones are in the Google Doc spreadsheet * Should have a version ready by the end of the week * Will have by Friday: ***** PLEASE FILL OUT ****** IPAC (Xiuqin, Gregory): * Activities in the Google Doc spreadsheet * Milestones are in the Google Doc spreadsheet * Should have a version ready by the end of the week * Gregory is writing the document describing the QA tooling (summary of the Princeton meeting) - Will focus on work to be done in the next 6-12 months - Should be ready in the next two weeks * Will have by Friday: ***** PLEASE FILL OUT ****** NCSA (Don, Margaret): * Developing a plan responding to so-far understood business needs (operations milestones, commissioning milestones) * By end-of-year expect to have a plan incorporating these, ready for integration with software teams’ plans * That integration can begin Jan 2nd at the earliest * Would like a better sense of the needs of the software teams while in construction (i.e., not just “what” is being asked for, but a better understanding of why to know how to respond to the need). * K-T has added initial high-level milestones in to rows 38-54 * Following initial integration of software groups’ plans, MJ will try to distill the high-level business needs to allow for facility<->payloads (development) plan integration. * NCSA plan is in OmniPlan * Will have by Friday: ***** PLEASE FILL OUT ****** * General clarifications: - Q: what to do when a delivery date of a deliverable is not currently known. A: for your initial plan, assume you’ll receive it by your need-by date. We’ll be adjusting these after integration. - Producing an internally self-consistent integrated plan is our first priority. But we’re aware that that plan is coming in over budget & schedule. For the DMLT F2F, we need to begin identifying potential descopes or changes that would reduce the cost or time to construct, but would have a bearable impact on science enabled by LSST. I saw that the TCAMs have already started doing that — if you haven’t already, make sure to involve your Project Scientists in this (best to have them fully tasked with it). The PS are the Product Owners (and you’re already busy working on the plan). Corrections/additions welcome! Regards, — Mario Juric, Large Synoptic Survey Telescope Data Management Subsystem Lead Web : http://lsst.org Phone : +1 609 933 1033 _______________________________________________ LSST-DMLT mailing list LSST-DMLT@listserv.lsstcorp.org https://listserv.lsstcorp.org/mailman/listinfo/lsst-dmlt